Skip to main content

Full text of "Naven, a survey of the problems suggested by a composite picture of the culture of a New Guinea tribe drawn from three points of view"

See other formats 


^^ 


\  rpr..  .  inc  urn  Tf 


^3J0 


!•#■ 


^' 


^       ^ 


-^'^!M\\Mi\\^ 


4s^ 


i^i 


:S 


\RY/9/-,         .,>^MIRRARYQr  A\^E  UNIVERV/^^^^         .v 


811  #      ^OAHVaail^  '%13DNVS01^'^       "^/mMNllJWV^  >' 


o 


^ ^^ 


,OFCALIFO%        ^ 

1    il^ 


K(f 


"^^Aavaaii-i^^     '^^;. 


^lOSANCElfj> 


^ 


A\\t  UNIVERS/^        ^>;10SANCE15X^ 


o^ 


# 


•5- 


\>^EUNIVERJ// 


;AjOPCAilFO/?^ 


^ 


^. 


c^ 


5  1  tr"  ^   -^ 


"<3UdllYJJU 

^OFCAllF0/?,( 


n 


,.,v>> 


■S> 


\ 


aMEUNIVER%. 


S  '•■^, 


Hr 


NAVEN 


N  A  V  E  N 

A  Survey  of  the  Problems  suggested 

by  a  Composite  Picture  of  the  Culture  of  a 

New  Guinea  Tribe  drawn  from 

Three  Points  of  View 


BY 

GREGORY  BATESON 


SECOND  EDITION 


STANFORD  UNIVERSITY  PRESS 

STANFORD,  CALIFORNIA 
1958 


STANFORD   UNIVERSITY  PRESS 

Stanford,  California 
London:  oxford  university  press 

First  edition  fublished  1936  - 
Second  edition  fublished  1958 

©1958  by  the  Board  of  Trustees  of  the 

Leland  Stanford  Junior  University 

All  rights  reserved 

Library  of  Cofigress  Catalog  Card  IS  umber:  ^8-8  j  20 
Printed  in  the  United  States  of  America 


'/l^ 


To 
DR  A.  C.  HADDON,  F.R.S. 


PREFACE  TO  THE  SECOND  EDITION 

IN  this  edition,  the  body  of  the  book  has  been  left  un- 
changed. In  it,  a  hne  of  theoretical  thought  is  pursued 
as  far  as  was  possible  in  1936.  But  problems  similar  to 
those  which  I  posed  and  left  unsolved  have  since  been  partly 
solved  by  cybernetics  and  communications  theory.  I  have 
therefore  added  to  the  book  a  second  epilogue,  in  which  the 
theoretical  positions  of  the  book  are  discussed  in  the  light  of 
modern  developments. 

What  has  happened  has  been  the  growth  of  a  new  way  of 
thinking  about  organization  and  disorganization.  Today,  data 
from  a  New  Guinea  tribe  and  the  superficially  very  different 
data  of  psychiatry  can  be  approached  in  terms  of  a  single  epis- 
temology— a  single  body  of  questions. 

We  now  have  the  beginnings  of  a  general  theory  of  process 
and  change,  adaptation  and  pathology;  and,  in  terms  of  the 
general  theory,  we  have  to  reexamine  all  that  we  thought  we 
knew  about  organisms,  societies,  famihes,  personal  relation- 
ships, ecological  systems,  servo-mechanisms,  and  the  like. 

GREGORY  BATESON 

Palo  Alto,  California 
January  1958 


FOREWORD 

THE  anthropologist  on  his  travels  incurs  many  debts. 
He  is  indebted  to  those  who  have  helped  him  financially, 
he  is  indebted  to  the  intellects  of  other  anthropologists, 
and  he  is  indebted  to  a  great  number  of  people  who  help  him 
both  materially  and  by  their  friendship  in  the  countries  to 
which  he  goes.  For  my  first  expedition^  I  have  already  ex- 
pressed my  indebtedness  to  those  who  helped  me,  and  for 
my  second  expedition  the  list  is  an  even  longer  one. 

For  the  financing  of  the  expedition  I  am  indebted  to  the 
Percy  Sladen  Memorial  Fund,  to  the  Royal  Society  Govern- 
ment Grants  Committee,  and  to  St  John's  College,  Cambridge, 
of  which  body  I  was  a  research  Fellow,  an  extraordinary 
degree  of  freedom  being  permitted  me  by  the  Council. 

Intellectually,  I  have  to  thank  first  and  foremost  my 
teachers,  Dr  A.  C.  Haddon,  Professor  A.  R.  Radcliflfe-Brown, 
and  Professor  Malinowski.  Dr  Haddon  first  made  me  an 
anthropologist,  telling  me  in  a  railway  train  between  Cam- 
bridge and  King's  Lynn  that  he  would  train  me  and  send  me 
to  New  Guinea.  Later  his  training  was  supplemented  by 
Professors  Radcliff'e-Brown  and  Malinowski,  and  to  all  of  these 
I  owe  much.  In  this  Foreword,  I  wish  to  stress  my  admira- 
tion for  Professor  Malinowski's  work.  In  the  body  of  the 
book  I  have  from  time  to  time  been  critical  of  his  views  and 
theoretical  approach.  But,  of  course,  I  recognise  the  im- 
portance of  his  contribution  to  anthropology,  and  though 
I  may  think  it  time  for  us  to  modify  our  theoretical  approach, 
the  new  theoretical  categories  which  I  advocate  are  largely 
built  upon  ideas  implicit  in  his  work. 

For  their  criticism  and  for  conversations  which  I  have  had 
with  them,  I  wish  to  thank  a  number  of  people  who  have 
read  parts  of  the  manuscript  of  this  book — Professor  Rad- 
cliffe-Brown,  Professor  F.  C.  Bartlett,  Dr  J.  T.  MacCurdy, 
Dr  Ruth  Benedict,  Dr  E.  J.  Lindgren,  Mr  John  Layard  and 
Mr  C.  H.  Waddington. 

^  Oceania,  1932,  vol.  11,  nos.  3  and  4. 


X  Foreword 

Of  those  who  helped  me  in  New  Guinea,  I  am  most 
indebted  to  Dr  Margaret  Mead  and  Dr  R.  F.  Fortune,  whom 
I  met  on  the  Sepik  River.  At  a  time  when  I  was  hopelessly 
sick  of  field  work,  I  had  the  good  fortune  to  meet  these 
workers  who  set  me  a  new  and  higher  standard  of  work  in 
the  field.  In  conversations  with  them  my  approach  to  anthro- 
pological problems  had  its  origin. 

While  we  were  all  three  of  us  working  on  the  Sepik  River, 
there  arrived  from  America  a  part  of  the  manuscript  of 
Dr  Benedict's  Patterns  of  Culture,  and  this  event  influenced 
my  thinking  very  profoundly.  For  practical  help  I  have  to 
thank  first  and  foremost  the  Government  of  the  mandated 
territory  of  New  Guinea,  especially  the  Government  anthro- 
pologist, Mr  E.  W.  P.  Chinnery,  Mr  E.  D.  Robinson,  the 
District  Officer  of  the  Sepik  district,  and  Patrol  Officers 
Keogh,  Beckett,  Thomas  and  Bloxham.  Opposition  of 
Government  officials  to  intrusive  anthropologists  would  be 
understandable,  but  I  have  not  experienced  it.  In  every 
contact  which  I  had  with  these  officials  they  were  my  friends, 
and  often  they  were  my  hosts. 

As  a  result  of  my  journey  to  New  Guinea  I  number  many 
friends  among  its  population:  Mr  and  Mrs  Mackenzie  of 
the  Lady  Betty,  Mr  and  Mrs  E.  J.  Wauchope,  Messrs  William 
and  Roy  McGregor,  Bob  Overall,  Bill  Mason,  Gibson  and 
Eichorn.  Last  but  not  least  I  have  to  thank  my  native  in- 
formants who  were  my  friends,  although  I  appeared  to  them 
ridiculous :  Mali-kindjin,Tshamelwan,Tshava,  Mbana,  Kaing- 
genwan,  Kambwankelabi,  and  my  cook-boy,  Momwaishi. 


GREGORY  BATESON 


St  John's  College,  Cambridge 
December  1935 


TABLE  OF  CONTENTS 


This  Table  is  provided  to  help  those  who  may  be  interested  in  dis- 
secting the  arguments  contained  in  this  book.  Every  heading  in  the 
book  is  listed  in  the  Table,  and,  in  addition,  a  number  of  finer  sub- 
divisions, not  specially  marked  in  the  text,  are  here  given  the  dignity  of 
headings. 


Preface  to  the  Second  Edition  vii 

Foreword  ix-x 

Chap.  I.   METHODS  OF  PRESENTATION  1-5 

Artistic  and  scientific  techniques  contrasted;  functional  analysis; 
the  importance  of  ethos ;  synchronic  study  of  culture ;  sketch  of  the 
latmul  people. 

Chap.  II.   THE  NAVEN  CEREMONIES  6-22 

Occasions  on  which  Naven  are  performed  6 

(i)  Major  achievements  of  laua;  (2)  first  accomplishment  of  minor 
acts;  (3)  acts  characteristic  of  laua;  (4)  boasting  in  presence  oi  wau; 
(5)  changes  in  lauas  social  status;  naven  for  girls. 

Materials  upon  which  the  Description  is  based  10 

Small  naven ;  the  five  naven  witnessed  by  the  writer ;  native  descrip- 
tions of  naven. 

Description  of  the  Ceremonies  12 

Two  waus  in  Palimbai  naven;  tcaus  called  "mothers";  their  trans- 
vestite  costume;  their  buffoonery;  they  search  for  laua;  grotesque 
obscenities;  laua  gives  valuables  to  wau;  wau  rubs  buttocks  on 
laua's  leg;  classificatory  spreading  of  naven;  naven  in  Mindimbit  for 
children  who  had  made  sago;  transvestite  women;  relatives  con- 
cerned; their  costume  contrasted  with  that  of  wau;  costumes  of 
mother  and  mother's  brother's  wife;  special  kinship  terms  for 
transvestite  women ;  women  beating  men ;  women  enter  ceremonial 
house;  women's  dance;  naven  for  girl  who  had  caught  fish;  wau 
carrying  laua ;  laua  on  belly  of  wau ;  mother's  brother's  wife  (mbora) 
dances ;  presentations  of  pigs ;  return  presents  of  valuables ;  naven 
for  homicide;  mbora  snatches  feather  ornament  from  iau;  mbora 
copulates  with  wau ;  laua  spears  the  fish-trap  and  steps  across  all  the 
women ;  summary  of  the  naven  behaviour  of  the  various  relatives. 


xii  Contents 

Chap.  III.   THE  CONCEPTS  OF  STRUCTURE  AND 
FUNCTION  23-34 

Structure  23 

Formulations  as  summaries  of  many  details  of  cultural  behaviour; 
"tradition"  equated  with  Structure;  definition  of  Cultural  Pre- 
mises ;  definition  of  Cultural  Structure ;  Social  Structure. 

Function  26 

Strict  and  popular  uses  of  the  term;  "useful"  functions;  institu- 
tions; classification  of  "functions"  is  a  necessary  preliminary  to 
definition  of  institutions ;  previous  classifications ;  Radcliffe-Brown's 
system ;  the  position  of  economics ;  definitions  of  categories  used  by 
the  writer;  considerations  leading  to  subdivision  of  "pragmatic 
function";  impossibility  of  guessing  at  affective  functions  until  the 
ethos  of  the  culture  is  kiiown ;  the  analogy  between  ethos  and  eidos ; 
relation  between  these  concepts  and  "Configuration";  standardisa- 
tion of  individuals  by  culture;  the  irrelevance  of  psychology  to 
sociology. 

Chap.  IV.   CULTURAL   PREMISES   RELEVANT   TO 
THE  WAU-LAUA  RELATIONSHIP  35-53 

Identification  35 

Discrimination  between  Wau  and  Father  36 

Story  of  children  who  are  ashamed  of  father  and  mother  who 
behave  like  zvau  and  mbora ;  esoteric  knowledge  given  to  son  but 
sold  to  laua;  "my  laua  will  help  my  son";  wau  may  help  laua  in 
first  killing,  but  father  must  not  do  this ;  wau  and  laua  identified  in 
achievements ;  father  and  son  identified  in  economics. 

Identification  between  Father  and  Son  38 

Evidence  from  terminology;  double  terms  for  groups  of  kin;  the 
term,  wau,  sometimes  applied  to  mother's  brother's  son ;  avoidances 
between  father  and  son;  promotion  of  son  into  father's  initiatory 
grade ;  intimacy  between  father  and  son  is  shocking ;  mutual  respect. 

The  Child's  Relationship  to  the  Maternal  Clan  42 

Bones  regarded  as  a  product  of  semen,  but  flesh  and  blood  derived 
from  menstrual  blood ;  names  given  to  child  by  paternal  and 
maternal  clans;  -awan  names;  aspects  of  child's  personality  repre- 
sented by  these  names ;  fate  of  names  after  death ;  the  laua  identified 
with  the  ancestors  of  the  maternal  clan;  laua  addressed  as  "father 
and  father's  father";  laua  addressed  as  totem  of  maternal  clan;  laua 
dances  in  masks  representing  maternal  ancestors;  laua  ornaments 
himself  with  totemic  plants  of  maternal  clan ;  laua  eats  the  sacrifices 
to  his  maternal  ancestors ;  laua  eats  fowl  presented  by  trespasser ; 
laua  eats  maternal  ancestors;  pwivu  ritual;  mintshanggu  mortuary 
ceremony ;  part  played  by  paternal  and  maternal  clans. 


Contents  xiii 

Chap.  IV.    Cultural  Premises  etc.  (cont.) 

Summary  of  the  First  Three  Premises  48 

The  child  is  identified  with  the  father  but  competes  with  him  in 
achievement;  the  child  is  the  achievement  of  the  mother  and  his 
achievements  are  her  achievements ;  achievements  of  man  and  woman 
stressed  in  mortuary  ceremonial;  myth  of  naven  celebrated  for  a 
man  on  occasion  of  the  birth  of  his  child ;  the  tragic  consequences ; 
procedure  at  first  kill  explained ;  achievements  equated  with  ancestors. 

Identification  between  Brother  and  Sister  49 

Evidence  from  the  naming  system ;  ceremonial  behaviour  expressing 
this  identification ;  the  passing  of  the  right  to  names  from  sister  to 
brother ;  brother  cutting  sister's  hair,  presenting  a  shell  valuable  and 
claiming  the  names ;  position  of  woman  who  is  solitary  surviver  of 
clan ;  her  bride  price  greater  because  all  the  names  are  vested  in  her. 

Identification  of  Wife  and  Husband  52 

Evidence  from  kinship  terminology ;  household  as  a  single  economic 
unit;  context  and  identification. 


Chap.  V.   SORCERY  AND  VENGEANCE  54-73 

The  native  concept  of  ngglambi;  infectious  guilt;  native  phrasings 
of  the  cause  of  death;  ngglambi  equated  with  lex  talionis;  homo- 
nymous uses  of  the  word  nggamhwa,  "  vengeance  " ;  native  phrasings 
of  the  cause  of  death  illustrate  the  identifications  described  in 
Chapter  iv ;  eleven  stories  of  sorcery  and  killing  documenting  these 
identifications. 


Chap.  VI.    STRUCTURAL  ANALYSIS  OF  THE  WAU- 
LAUA  RELATIONSHIP  74-85 

The  wau's  behaviour  analysed  in  terms  of  the  identifications  de- 
scribed in  Chapter  iv  and  documented  in  Chapter  v ;  these  identifi- 
cations indicate  that  wau  is  a  "mother"  and  a  "brother-in-law"  of 
laua. 

I.    The  IF«w  as  a  " Mother "  75 

Sketch  of  the  relationship  between  a  mother  and  her  child ;  her 
gifts  of  food  to  the  child;  her  vicarious  pride;  the  mother  as  a 
comforter ;  analogous  behaviour  of  wau ;  exaggeration  in  wau  of  the 
natural  behaviour  of  mother;  behaviour  of  laua  analogous  to 
behaviour  of  son  towards  mother ;  his  loyalty. 


xiv  Contents 

Chap.  VI.    Structural  Analysis  etc.  (cont.) 

2.  The  Wau  as  a  "Wife's  Brother"  78 

Sketch  of  the  relationship  beuveen  brothers-in-law;  bride-price; 
indebtedness ;  co-operation ;  joking  about  mutual  opposition ;  in- 
sistence upon  loyalrj'  bet\veen  brothers-in-law ;  dut>'  of  peace-making 
between  maternal  clan  and  wife's  clan;  analogies  between  wau-laua 
relationship  and  that  of  brothers-in-law ;  laua  presents  valuables  to 
wau. 

3.  Other  Details  of  Wau&  Behaviour  82 

The  ceremonial  in  which  wau  rubs  his  buttocks  on  laua's  leg  is  not 
describable  in  terms  of  either  aspect  of  the  wau's  position;  com- 
bination of  identifications  would  label  wau  as  "wife"  of  laua; 
suggestion  that  the  ceremonial  is  an  expression  of  this  relationship ; 
supporting  evidence  for  this;  problem  of  elaboration  of  culture; 
details  of  behaviour  which  might  be  described  in  terms  of  weak 
identification  between  wau  and  laua's  father;  exaggeration  of  wau's 
behaviour;  the  relationship  between  wau  and  laua  summarised  in 
a  diagram. 


Chap.  VH.   THE  SOCIOLOGY  OF  NAVEN        86-107 

The  Integration  of  latmul  Communities  86 

Assumption  that  by  naven  the  kinship  link  between  wau  and  laua 
is  strengthened ;  the  importance  of  statistics  in  Sociology ;  the  type 
of  statistical  information  required ;  an  outline  of  the  marriage  system 
of  the  latmul,  showing  that  in  this  society  there  is  no  mechanism 
for  the  repetition  of  marriages  in  succeeding  generations;  iai 
marriage;  marriage  with  father's  sister's  daughter;  exchange  of 
women ;  logical  relationships  between  these  discordant  types  of 
marriage ;  irregular  marriages ;  the  importance  of  affinal  links  in  the 
integration  of  latmul  society;  the  behaviour  which  accompanies 
these  links ;  two  t^^jes  of  extended  affinal  relationship ;  lanoa  nampa 
and  laua  nyanggu;  laua  nyanggu  defined  by  past  marriages;  "women 
hither,  women  thither";  stressing  laua  relationship  equivalent  to  a 
stressing  of  old  affinal  links ;  size  of  community  limited  by  internal 
cohesion ;  fission  follows  patrilineal  linkages  and  breaks  the  affinal 
links;  the  weakness  of  the  latter  therefore  sets  limit  to  the  size  of 
community. 

Peripheral  and  Centripetal  Systems  97 

Analogous  methods  of  integrating  societies;  the  social  function  of 
codified  law  and  established  authority;  absence  of  such  mechanisms 
among  the  latmul;  system  of  sanctions  among  the  latmul;  the 
vengeance  sanction ;  quarrels  are  always  between  peripheral  groups, 
never  between  higher  and  lower  authority;  four  cases  illustrating 
this ;  a  thief  killed ;  a  woman  caught  spying  on  secret  flutes ;  a  junior 
ceremonial  house  defiled ;  a  quarrel  about  suspected  adultery. 


Contents  xv 

Chap.  VII.    The  Sociology  of  Naven  (cont.) 

Contrasting  types  of  Fission  io6 

Fission  of  latmul  communities  with  peripheral  orientation  leads  to 
formation  of  new  communities  with  the  same  cultural  norms  as  the 
parent ;  fission  of  European  systems  with  centripetal  organisation 
leads  to  formation  of  daughter  groups  with  divergent  norms. 

Chap.  VIII.   PROBLEMS  AND  METHODS  OF 

APPROACH  108-122 

Problems  108 

Structural  and  Sociological  analysis  has  answered  a  number  of 
questions  about  naven;  other  questions  remain;  exaggeration  of 
zvau's  behaviour;  problem  of  the  size  of  villages;  problems  of 
motivation;  wau's  hypothetical  desire  for  allegiance;  answers  based 
upon  hypothetical  "human  nature";  difficulties  in  attributing 
affective  motive. 

Zeitgeist  and  Configuration  1 1 1 

The  historian's  approach  to  culture ;  Zeitgeist  and  cultural  change ; 
Configuration  and  the  adoption  of  foreign  traits  of  culture ;  cultural 
emphases,  due  to  standardisation  of  individuals;  standardisation 
either  by  selection  or  training. 

Psychological  Theories  and  Ethology  114 

Criticism  of  answers  which  invoke  universal  human  nature ;  existence 
of  opposite  tendencies  in  human  nature ;  need  for  a  criterion  which 
shall  justify  us  in  invoking  one  tendency  rather  than  another ;  concept 
of  standardisation  provides  this  criterion ;  we  must  verify  that  the 
sentiments  invoked  are  actually  fostered  in  the  culture;  circular 
argument;  its  justification;  definition  of  ethos  (p.  118);  ethos  and 
typology;  possibility  that  future  comparative  work  will  provide 
verification  of  ethological  hypotheses. 

Examples  of  Ethos  in  English  Culture  119 


Chap.  IX.   THE    ETHOS    OF    lATMUL    CULTURE: 
THE  MEN  123-141 

The  Ceremonial  House  123 

The  ceremonial  house  compared  with  a  church;  behaviour  in  the 
ceremonial  house ;  self-consciousness ;  debates ;  pride  in  totemic 
ancestors ;  stealing  of  names  and  ancestors ;  ritual  staged  for  women. 


xvi  Contents 

Chap.  IX.    The  Ethos  of  latmul  Culture  etc.  {cont.) 

Initiation  129 

Irresponsible  bullying ;  scarification ;  hazing  of  novices ;  competition 
between  the  moieties;  novices  as  "wives"  of  the  initiators;  etho- 
logical  processes  in  initiation;  "cutting  off  their  own  noses  to  spite 
the  other  fellow's  face  " ;  a  woman  sees  a  whistle  in  Mindimbit,  and 
the  secrets  are  therefore  shown  to  the  small  boys;  small  boy  killed 
in  Palimbai  for  insulting  wagan,  and  the  wagan  are  therefore  shown 
to  women. 

Headhunting  137 

A  captive  speared;  feuds;  personal  pride  and  village  prosperity; 
inability  to  revenge  causes  ngglambi;  enemy  corpse  ritually  killed; 
heads  and  phallic  standing  stones;  the  vanquished  give  the  names 
of  the  dead. 


Chap.X.   THE    ETHOS    OF    lATMUL    CULTURE: 
THE  WOMEN  142-151 

The  dwelling  house ;  fishing ;  markets ;  assertive  women ;  taking  the 
initiative  in  love;  women  in  headhunting;  seeking  vengeance; 
women's  courage  commemorated;  woman's  authority  in  the  house; 
double  emphasis  in  women's  ethos ;  the  same  double  emphasis  in 
ceremonial;  jolly  dances  for  women  only;  innocent  obscenities; 
women  proud  in  public  procession ;  mild  transvesticism. 


Chap.  XI.   ATTITUDES  TOWARDS  DEATH    152-159 

A  death  at  night;  the  women  weep;  a  man  is  embarrassed;  the 
burial ;  the  death  of  a  great  fighter ;  the  men  debate ;  they  set  up  a 
figure  of  the  dead  with  s>Tnbols  of  his  achievements ;  death  provides 
a  contest  for  competitive  boasting;  later  mortuary  ceremonies; 
mifitshanggu;  "quiet  singing";  women's  dirges  stimulate  the  men 
to  caricature ;  pride  in  the  presence  of  certain  death. 


Chap.  XII.   THE  PREFERRED  TYPES  160-170 

Ethological  contrast  and  Kretschmer's  typology;  the  violent  man 
and  the  man  of  discretion;  a  "cranky"  informant;  types  contrasted 
in  mythology;  long  noses;  typology  and  phallic  symbolism;  Mali- 
kindjin,  a  character  sketch;  ambivalent  feelings  about  him;  skinny 
sorcerers ;  native  personality  and  culture  contact ;  Tshimbat,  a 
maladjusted  individual ;  his  pig  killed. 


Contents  xvii 

Chap.  XIII.   ETHOLOGICAL  CONTRAST,  COMPETI- 
TION AND  SCHISMOGENESIS  171-197 

Heredity  and  Environment  lyi 

Biological  differences  between  the  sexes ;  difficulty  of  accounting  for 
ethological  contrast  in  these  terms;  the  possibility  that  genetic 
variation  has  provided  clues  for  cultural  emphasis. 

Conditioning  factors  which  maintain  Sex  Contrast        173 

Possibility  that  male  ethos  is  inculcated  in  training  of  youths; 
imitation  of  seniors ;  headhunting ;  first  kill ;  such  factors  maintain 
the  status  quo. 

Schismogenesis  175 

The  status  quo  as  a  dynamic  equilibrium;  schismogenesis  defined; 
tendencies  towards  progressive  change  in  behaviour  patterns  in 
relationships ;  complementary  and  symmetrical  schismogenesis. 

Schismogenesis  in  latmul  Culture  177 

Women  as  an  audience;  reaction  of  the  men  to  women's  dirging; 
complementary  schismogenesis  in  initiation;  the  boasting  of  the 
laua;  symmetrical  schismogenesis  in  initiation. 

Schismogenesis  in  Other  Contexts  178 

(i)  In  marriage;  (2)  in  progressive  psychological  maladjustment; 
schismogenesis  within  the  personality;  the  importance  of  cultural 
emphases;  (3)  in  culture  contacts;  (4)  in  politics. 

The  Progress  of  Schismogenesis  187 

Schismogenic  behaviour  at  first  satisfactory;  subsequent  distortion 
of  personalities;  mutual  hostility;  mutual  envy;  distaste  for  com- 
plementary ethos;  schismogenesis  expected  to  follow  exponential 
curve ;  two  factors  which  may  hasten  schismogenesis. 

The  Control  of  Schismogenesis  190 

The  nature  of  dynamic  equilibrium;  factors  which  preserve  the 
status  quo;  upper  limits  of  tolerance;  over-emphasis  of  ethos; 
processes  of  change  opposed  to  schismogenesis;  (i)  complementary 
patterns  in  a  symmetrical  relationship  and  symmetrical  patterns  in 
a  complementary  relationship ;  (2)  schismogenesis  based  upon  one 
pair  of  complementary  patterns  may  be  restrained  by  patterns  of  a 
different  complementary  pair;  (3)  sudden  change  in  the  patterns  of 
a  symmetrical  schismogenesis ;  (4)  schismogenesis  between  two 
groups  may  be  checked  by  schismogenic  relationship  with  an  outside 
group;  (5)  balanced  hierarchies,  etc.;  (6)  conscious  control  of 
schismogenesis;  complicated  latmul  mechanisms;  (7)  mutual  de- 
pendance  between  conflicting  groups;  (8)  progressive  changes  in 
behaviour  resulting  in  convergence. 


xviii  Contents 

Chap.  XIV.   THE  EXPRESSION  OF  ETHOS  IN  A^^  VEN 

198-217 

Sex  Ethos  and  Naven  198 

Many  of  the  details  of  naven  now  appear  as  ethologically  normal; 
exaggerated  behaviour  of  wau ;  mother  lying  naked ;  women  showing 
pride  in  public  ceremonial ;  problems  of  transvesticism ;  the  fashion- 
able horsewoman ;  a  theory  of  latmul  transvesticism. 

Kinship  Motivation  and  Naven  203 

Wau  (p.  203);  his  behaviour  interpretable  in  structural  terms;  the 
incompleteness  of  such  interpretation;  factitious  emotion;  two 
degrees  of  falsity  in  the  emotions  attributed  to  the  classificatory  wau ; 
myths  which  indicate  some  hostility  between  wau  and  laua ;  analysis 
of  this  hostility  in  terms  of  wau's  structural  position  as  defined  by 
identifications;  hostility  connected  with  wau's  position  as  "brother- 
in-law";  effects  of  this  on  his  relationship  with  laua;  the  wau 
symbolically  claiming  the  achievements  of  his  laua. 

tazvontu  (p.  210);  differences  and  analogies  between  his  position 
and  that  of  ivau ;  his  overt  expression  of  hostility. 

nyai',  nyamun  and  tshuambo  (p.  210) ;  sociological,  economic,  struc- 
tural and  emotional  factors  which  prevent  the  father  from  taking  an 
active  part  in  naven ;  analogous  factors  in  the  position  of  brothers. 

Contrast  between  the  motivation  of  the  men  and  that  of  the  women 
(p.  213). 

nyame  (p.  215);  a  straightforward  expression  of  negative  self- 
feeling  ;  nakedness  in  various  contexts — mourning,  supplication  and 
rejoicing. 

iau  (p.  215);  her  identification  with  the  father  would  provide 
structural  basis  for  two  opposed  behaviour  patterns;  she  adopts 
swaggering  role. 

tshaishi  (p.  216);  analogies  with  iau;  levirate. 

nyanggai  (p.  216);  her  quarrel  with  tshaishi. 

mbora  (p.  217);  her  identification  with  transvestite  wau  results  in 
ambiguity;  snatching  the  feather  ornament  from  iau,  symbolically 
claiming  the  achievement. 


Chap.  XV.   THE  EIDOS  OF  lATMUL  CULTURE 

218-256 

The  Problem  of  latmul  Complexity  218 

The  nature  of  cultural  structure;  the  role  of  the  scientist;  eidos 
defined  (p.  220);  standardisation  of  intellectual  activity;  selection  or 
training;  spells  to  improve  memory. 

Intellectual  Activation  in  latmul  Culture  221 

Memory;  erudition  in  debate;  rote  memory  probably  unimportant; 
secret  mythology  handled  as  a  series  of  details  rather  than  as  nar- 
rative; visual  and  kinaesthetic  imagery;  naven  as  an  expression  of 
eidos. 


Contents  xix 

Chap.  XV.    The  Eidos  of  latmul  Culture  {cont.) 

The  Pervasive  Nature  of  Eidos  226 

Activation  of  memory  not  confined  to  special  individuals ;  masters 
of  ceremony  moulding  the  culture;  the  naming  system  not  an 
isolated  context ;  the  whole  culture  reflected  in  the  system ;  initiation 
the  only  exception. 

latmul  Paradoxes  229 

Day  and  Night ;  the  new  Sun  and  the  old  Sun ;  ripples  and  waves ; 
photographic  development  as  a  clan  secret;  the  Aiitipodes;  lunar 
eclipse ;  confusion  of  sociological  and  affective  truth  with  cognitive 
reality ;  discrimination  and  identification  of  aspects  of  personality ; 
wagan;  borassus  palm  is  a  fish;  the  same  paradoxes  in  kinship; 
patriliny  versus  matriliny ;  the  same  paradoxes  in  naven. 

Other  Patterns  of  latmul  Thought  235 

Pluralism ;  monism ;  direct  and  diagonal  duahsm ;  types  of  dualism 
and  schismogenesis  (fn.  2,  p.  237) ;  "  artificial "  brothers  and  brothers- 
in-law  ;  kaishi;  other  types  of  dualism ;  confusion  between  direct  and 
diagonal ;  dualisms  and  age  differences ;  difference  in  sex  equivalent 
to  difference  in  age ;  staggered  series ;  generations ;  initiatory  grades ; 
brothers;  flutes;  Castor  and  Pollux  and  comparative  methods; 
latmul  and  Australian  kinship  compared. 

Scientific  and  Native  Ways  of  Thought  250 

Sociological  thinking  among  the  latmul;  structural  thinking;  etho- 
logical  thinking ;  economic  thinking ;  thinking  in  terms  of  character 
formation;  diachronic  and  synchronic  thinking. 

Other  Types  of  Psychological  Standardisation  255 

Apollonian  and  Dionysian;  tempo;  perseveration. 

Chap.  XVI.   EPILOGUE  1936  257-279 

Narrative  of  the  writer's  analysis  of  methods ;  lack  of  orientation  in 
field  work;  "Patterns  of  Culture";  adding  ethos  to  naven;  trans- 
vesticism ;  the  fallacy  of  misplaced  concreteness ;  structure  not  a  part 
but  an  aspect  of  culture ;  tendency  to  confuse  aspects ;  the  isolation 
of  sociology;  the  separation  of  ethos  from  eidos;  configuration; 
schismogenesis;  practical  order  of  methods  of  approach;  various 
views  of  schismogenesis;  the  equivalence  between  the  two  types 
of  dualism  and  the  two  types  of  schismogenesis ;  the  inculcation  of 
eidos;  "affective"  and  "cognitive"  aspects  of  personality  defined 
in  terms  of  stimulus-response  events;  various  methods  of  approach 
introduce  various  distortions  into  the  picture  of  schismogenesis; 
evaluation  of  methods  and  the  importance  of  separating  them. 

Chaf.  XVII.   EPILOGUE  1958  280-303 

Index  Glossary  of  Technical  and  Native  Words         307-312 


ILLUSTRATIONS 

PLATES 
Plates  Following  page  3 1 2 

TEXT  FIGURES 

Fig.  I.    The  wau's  kinship  position  84 

Fig.  2.    Marriage    with    the    father's    sister's 

daughter  89 

Fig.  3.    Kinship  traced  through  women  94 

Fig.  4,    Diagram  of  initiatory  groups  245 

Fig.  5.    The  study  of  the  cognitive  aspects  of 

personahty  275 

Fig.  6.    The  study  of  the  affective  aspects  of 

personahty  275 

Diagram  of  the  Kinship  System  305 


NAVEN 


CHAPTER  I 

Methods  of  Presentation 

IF  it  were  possible  adequately  to  present  the  whole  of  a 
culture,  stressing  every  aspect  exactly  as  it  is  stressed  in 
the  culture  itself,  no  single  detail  would  appear  bizarre  or 
strange  or  arbitrary  to  the  reader,  but  rather  the  details  would 
all  appear  natural  and  reasonable  as  they  do  to  the  natives 
who  have  lived  all  their  lives  within  the  culture.  Such  an 
exposition  may  be  attempted  by  either  of  two  methods,  by 
either  scientific  or  artistic  techniques.  On  the  artistic  side 
we  have  the  works  of  a  small  handful  of  men  who  have  been 
not  only  great  travellers  and  observers  but  also  sensitive 
writers — such  men  as  Charles  Doughty;  and  we  have  also 
splendid  representations  of  our  own  culture  in  such  novels 
as  those  of  Jane  Austen  or  John  Galsworthy,  On  the  scien- 
tific side  we  have  detailed  monumental  monographs  on  a  few 
peoples ;  and  recently  the  works  of  Radcliffe-Brown,  Malin- 
owski  and  the  Functional  School. 

These  students  have  set  themselves  the  same  great  task, 
that  of  describing  culture  as  a  whole  in  such  a  manner  that 
each  detail  shall  appear  as  the  natural  consequence  of  the 
remainder  of  the  culture.  But  their  method  differs  from  that 
of  the  great  artists  in  one  fundamental  point.  The  artist  is 
content  to  describe  culture  in  such  a  manner  that  many  of 
its  premises  and  the  inter-relations  of  its  parts  are  implicit 
in  his  composition.  He  can  leave  a  great  many  of  the  most 
fundamental  aspects  of  culture  to  be  picked  up,  not  from  his 
actual  words,  but  from  his  emphasis.  He  can  choose  words 
whose  very  sound  is  more  significant  than  their  dictionary 
meaning  and  he  can  so  group  and  stress  them  that  the  reader 
almost  unconsciously  receives  information  which  is  not  ex- 
plicit in  the  sentences  and  which  the  artist  would  find  it 
hard — almost  impossible — to  express  in  analytic  terms.  This 
impressionistic  technique  is  utterly  foreign  to  the  methods  of 
science,  and  the  Functional  School  have  set  out  to  describe 


2  Methods  of  Presentation 

in  analytic,  cognitive  terms  the  whole  interlocking — almost 
living — nexus  which  is  a  culture. 

Very  properly  and  naturally  they  have  paid  greatest  atten- 
tion to  those  aspects  of  culture  which  lend  themselves  most 
readily  to  description  in  analytic  terms.  They  have  described 
the  structure  of  several  societies  and  shown  the  main  outlines 
of  the  pragmatic  functioning  of  this  structure.  But  they  have 
scarcely  attempted  the  delineation  of  those  aspects  of  cul- 
ture which  the  artist  is  able  to  express  by  impressionistic 
methods.  If  we  read  Arabia  Deserta,  we  are  struck  by  the 
astonishing  way  in  which  every  incident  is  informed  with  the 
emotional  tone  of  Arab  life.  More  than  this,  many  of  the 
incidents  would  be  impossible  with  a  different  emotional 
/background.  Evidently  then  the  emotional  background  is 
causally  active  within  a  culture,  and  no  functional  study  can 
•  ever  be  reasonably  complete  unless  it  links  up  the  structure 
'  and  pragmatic  working  of  the  culture  with  its  emotional  tone 
i  or  ethos.i 

The  present  work  is  a  description  of  certain  ceremonial 
behaviour  of  the  latmul  people  of  New  Guinea  in  which  men 
dress  as  women  and  women  dress  as  men,  and  an  attempt — 
crude  and  imperfect,  since  the  technique  is  new — to  relate 
this  behaviour,  not  only  to  the  structure  and  pragmatic 
functioning  of  latmul  culture,  but  also  to  its  ethos. 

This  enquiry  will  involve  me  in  a  perhaps  tedious  discussion 
of  abstractions,  and  to  those  who  have  some  difficulty  with, 
or  a  distaste  for,  epistemology  I  would  recommend  that  they 
read  first  the  descriptive  chapters,  especially  those  on  the 
ethos  of  latmul  culture,  so  that  a  preliminary  study  of  con- 
crete examples  may  make  clear  my  abstractions.  Others  may 
find  that  the  Epilogue,  in  which  I  have  recounted  some  of  the 

^  At  a  later  stage  I  shall  examine  and  attempt  to  define  these  concepts 
of  Structure,  Function  and  Ethos  in  a  more  critical  manner,  but  since 
the  concept  of  Ethos  is  still  unfamiliar  to  many  anthropologists,  it  may 
be  well  to  insert  here  the  definition  which  the  Oxford  English  Dictionary 
gives  of  the  word.  The  definition  reads:  "i.  The  characteristic  spirit, 
prevalent  tone  of  sentiment  of  a  people  or  community;  the  'genius'  of 
an  institution  or  system."  The  earliest  example  of  the  use  of  ethos  in 
this  sense  is  quoted  from  Palgrave,  1851.  I  must  ask  classical  scholars 
to  forgive  the  plural  "ethoses"  which  I  have  coined  in  preference  to 
"ethe".   The  latter  could  never  be  English. 


Methods  of  Presentation  3 

theoretical  errors  into  which  I  fell  in  the  course  of  my 
enquiry,  will  help  them  to  understand  my  present  theoretical 
position. 

At  this  early  stage,  I  wish  to  make  it  perfectly  clear  that 
I  do  not  regard  Ritual,  Structure,  Pragmatic  Functioning 
and  Ethos  as  independent  entities  but  as  fundamentally  in- 
separable aspects  of  culture.  Since,  however,  it  is  impossible 
to  present  the  whole  of  a  culture  simultaneously  in  a  single 
flash,  I  must  begin  at  some  arbitrarily  chosen  point  in  the 
analysis;  and  since  words  must  necessarily  be  arranged  in 
lines,  I  must  present  the  culture,  which  like  all  other  cultures 
is  really  an  elaborate  reticulum  of  interlocking  cause  and 
effect,  not  with  a  network  of  words  but  with  words  in  linear 
series.  The  order  in  which  such  a  description  is  arranged 
is  necessarily  arbitrary  and  artificial,  and  I  shall  therefore 
choose  that  arrangement  which  will  bring  my  methods  of 
approach  into  the  sharpest  relief.  I  shall  first  present  the 
ceremonial  behaviour,  torn  from  its  context  so  that  it 
appears  bizarre  and  nonsensical;  and  I  shall  then  describe 
the  various  aspects  of  its  cultural  setting  and  indicate  how 
the  ceremonial  can  be  related  to  the  various  aspects  of  the 
culture. 

Throughout  this  analysis  I  shall  confine  myself  to  syn- 
chronic explanations  of  the  phenomena,  that  is  to  say,  to 
explanations  which  invoke  only  such  other  phenomena  as  are 
now  present  in  latmul  culture,  or  which,  like  head-hunting, 
have  disappeared  so  recently  that  they  can  be  considered  as 
part  of  the  natural  setting  of  the  ceremonies  as  I  collected 
them.  I  shall  not  enquire  what  either  the  ceremonies  or  their 
cultural  setting  may  have  been  like  in  the  past.  In  my  use 
of  causal  terminology  I  shall  be  referring  to  conditional  rather 
than  to  precipitating  causes.  Thus  in  a  synchronic  study  of 
a  fire  I  should  say  that  the  fire  burns  because  there  is  oxygen 
in  the  room,^  etc.,  but  I  should  not  enquire  how  the  fire  was 
first  ignited. 

^  I  have  endeavoured  to  avoid  phrasings  reminiscent  of  the  statement, 
"the  square  of  the  velocity  is  a  cause  of  the  kinetic  energy".  But  when 
I  was  first  working  out  the  lines  of  thought  upon  which  this  book  is  based 
I  not  infrequently  fell  into  this  verbal  error  and  it  is  likely  that  in  some 
cases  the  mistake  has  passed  uncorrected. 


4  Methods  of  Presentation 

It  is  of  course  true  that  the  ceremonies  have  had  a  history 
and  it  would  no  doubt  be  possible  to  speculate  about  that 
history.  But  such  is  not  my  purpose.  I  shall  be  content  only 
to  show  some  of  the  sorts  of  functional  relationship  which 
exist  between  the  ceremonies  and  the  remainder  of  the  con- 
temporary culture  of  the  latmul.  Perhaps,  in  the  future,  a 
clearer  understanding  of  the  synchronic  aspects  of  society  will 
enable  us  to  isolate  and  define  the  diachronic,  the  processes 
of  cultural  change. 

The  latmul  people^  live  on  the  middle  reaches  of  the  Sepik 
River  in  the  Mandated  Territory  of  New  Guinea.  They  are  j 
a  fine,  proud,  head-hunting  people  who  live  in  big  villages 
with  a  population  of  between  two  hundred  and  a  thousand 
individuals  in  each  village.  Their  social  organisation,  kinship 
and  religious  systems  are  developed  to  an  extreme  of  com- 
plexity. The  community  is  subdivided  into  groups  according 
to  two  independent  systems  with  very  little  congruence  be- 
tween one  system  and  the  other.  On  the  one  hand  there  is  a 
division  into  two  totemic  moieties  which  are  further  sub- 
divided into  phratries  and  clans ;  and  on  the  other  hand  there 
is  a  division  into  two  cross-cutting  pairs  of  initiatory  moieties 
which  are  subdivided  into  age  grades.  None  of  these  groups 
are  strictly  exogamous.  Membership  of  all  the  groups  is 
determined  by  patrilineal  descent. 

In  spite  of  this  strong  emphasis  upon  patriliny,  the  people 
pay  a  great  deal  of  attention  to  kinship  links  through  the 
mother  or  sister,  and  both  the  patrilineal  and  matrilineal  links 
are  preserved  in  a  classificatory  system  through  many  genera- 
tions. Thus  the  term  wau  (mother's  brother)  and  laua  (sister's 
child,  m.s.)  are  used  not  only  between  own  mother's  brother 
and  sister's  children,  but  also  in  a  classificatory  way,  so  that 
the  term  wau  includes  such  relatives  as  mother's  mother's 
sister's  son  (cf.  Fig.  3,  p.  94)  even  though  all  three  of  the 

^  A  preliminary  account  of  this  people,  based  on  the  results  of  my  first 
expedition,  has  been  published  in  Oceania  (1932,  vol.  11,  nos.  3  and  4). 
The  present  paper  is  based  upon  the  work  of  this  first  expedition  as  well 
as  on  my  second  expedition  when  I  spent  a  further  fifteen  months  among 
the  latmul.  My  first  expedition  was  financed  by  Cambridge  University 
with  a  grant  from  the  Anthony  Wilkin  Fund.  The  second  was  financed 
jointly  by  the  Royal  Society  and  the  Percy  Sladen  Trust. 


Methods  of  Presentation  5 

Intervening  women  through  whom  the  kinship  is  traced  have 
married  into  different  clans.  It  is  with  the  relations  between 
classificatory  waus}-  and  lauas  that  this  book  chiefly  deals. 

^  In  the  latmul  language  there  is  no  suffix  attached  to  substantives  to 
indicate  plurality.  I  have  therefore  used  the  English  suffix,  -s,  for  this 
purpose.  In  place  of  the  latmul  possessive  suffix,  -na,  I  have  used  the 
ordinary  English  -'s.  Native  words  are  in  all  cases  italicised  but  the 
English  suffixes  remain  in  ordinary  type. 


CHAPTER  II 

The  Naven  Ceremonies 

OCCASIONS   ON  WHICH   NAVEN 
ARE   PERFORMED 

TH  E  ceremonies  are  called  naven  and  are  performed  in 
celebration  of  the  acts  and  achievements  of  the  laua 
(sister's  child).  Whenever  a  laua — boy  or  girl,  man  or 
woman — performs  some  standard  cultural  act,  and  especially 
when  the  child  performs  this  act  for  the  first  time  in  its  life, 
the  occasion  may  be  celebrated  by  its  ivau.  The  possible  occa- 
sions for  the  performance  of  naven  are  very  numerous  and 
very  frequent.  In  the  case  of  a  boy  the  list  of  acts  and  achieve- 
ments which  may  be  celebrated  in  this  way  is  a  long  one,  and 
may  conveniently  be  divided  into  five  categories : 

I .  Major  achievements  which,  though  greeted  with  a  more 
elaborate  naven  upon  their  first  performance,  are  also  received 
with  some  show  of  naven  behaviour  every  time  they  occur. 
Of  these  the  most  important  is  homicide.  The  first  time  a  boy 
kills  an  enemy  or  a  foreigner  or  some  bought  victim  is  made 
the  occasion  for  the  most  complete  naven,  involving  the 
greatest  number  of  relatives  and  the  greatest  variety  of  ritual 
incidents.  Later  in  his  life  when  the  achievement  is  repeated, 
there  will  still  be  some  naven  performance  on  the  part  of  the 
wau,  but  the  majority  of  the  ritual  incidents  will  probably 
be  omitted.  Next  to  actual  homicide,  the  most  honoured 
acts  are  those  which  help  others  to  successful  killing.  The 
man  standing  on  the  bow  of  a  war  canoe  does  not  carry  a 
spear-thrower  but  carries  instead  a  very  light  paddle  set  in 
a  long  bamboo  shaft;  with  this  he  wards  off  darts  from  the 
spear-throwers  of  the  enemy.  This  man  may  be  honoured 
with  naven  for  any  kills  effected  by  members  of  his  canoe. 
Another  act  contributory  to  killing  which  may  be  honoured  is 
the  enticing  of  foreigners  into  the  village  so  that  others  may 
kill  them.  Far  behind  these  in  importance  come  such  achieve- 
ments as  the  killing  of  a  large  crocodile,  the  killing  of  a  wild 


The  Naven  Ceremonies  7 

pig,  the  spearing  of  a  giant  eel,  etc. — achievements  which  are 
still  sufficiently  important  whenever  they  are  repeated  to 
stimulate  the  wau  at  least  to  a  ritual  greeting  and  perhaps  to 
the  throwing  of  a  cloud  of  lime  at  the  laua. 

2.  Minor  cultural  acts  which  are  celebrated  only  upon 
the  first  occasion  of  their  achievement.  Actually  it  would 
perhaps  be  correct  to  say  that  the  first  performance  of  any 
cultural  act  may  be  the  excuse  for  a  naven.  An  informant  in 
Mindimbit  village,  however,  gave  me  a  long  list,  and  I  have 
added  to  it  one  or  two  other  acts  which  I  know  may  be  cele- 
brated with  naven.  This  list,  though  of  course  it  is  not 
complete,  is  yet  worth  reproducing,  because  it  illustrates  the 
sort  of  act  which  may  be  noticed  by  the  wau.  It  includes: 
killing  any  of  the  following  animals — birds,  fish,  eel,  tortoise, 
flying  fox;  planting  any  of  the  following  plants — yams, 
tobacco,  taro,  coconut,  areca,  betel,  sago,  sugar-cane ;  spotting 
an  opossum  in  the  bush;  felling  a  sago  palm,  opening  it 
and  beating  sago;  using  a  spear-thrower;  using  a  throwing 
stick  to  kill  a  bird;  using  a  stone  axe  (or  nowadays  a  steel 
knife  or  axe);  sharpening  a  fish  spear;  cutting  a  paddle; 
making  a  canoe;  making  a  digging  stick;  making  a  spear- 
thrower;  incising  patterns  on  a  lime  gourd;  plaiting  an  arm 
band;  making  a  shell  girdle;  belting  a  hand  drum;  beating 
a  slit-gong ;  blowing  a  trumpet ;  playing  a  flute ;  beating  the 

^  secreT^it-gongs  called  wagan ;  travelling  to  another  village 
and  returning;  acquiring  shell  valuables;  buying  an  axe, 
knife,  mirror,  etc. ;  buying  areca  nut ;  killing  a  pig  and  standing 
a  feast. 

3.  Acts  characteristic  of  laua.  The  relationship  between 
laua  and  wau  is  marked  by  various  forms  of  characteristic 
behaviour — acts  which  each  performs  in  reference  to  the 
other.  Such  acts  may  be  described  as  duties,  services,  or 
privileges,  and  it  is  not  at  all  easy  in  any  given  case  to  decide 
which  of  these  terms  should  be  applied.  In  general,  whenever 
the  laua  performs  any  conspicuous  act  which  is  characteristic 
of  his  position  as  laua,  the  wau  will  respond  with  some  show 
of  naven  behaviour.  These  acts  include  such  ceremonials  as 
the  exhibition  of  totemic  ancestors  of  wau's  clan,  dancing 
in  masks  which  represent  these  ancestors  and  carving  their 


8  The  Naven  Ceremonies 

images  on  the  posts  which  are  to  stand  in  the  section  of  the 
ceremonial  house  which  belongs  to  the  wau's  clan.  Several 
of  the  musical  activities  mentioned  above — e.g.  beating  secret 
slit-gongs  and  playing  flutes,  may  also  fall  within  this  category 
of  services  to  the  wau's,  clan,  A  man  may  blow  the  flutes  of 
his  own  clan  and  his  musical  accomplishment  be  greeted  by 
naven  upon  the  first  occasion,  but  it  is  also  his  duty  or  privi- 
lege to  blow  the  totemic  flutes  of  his  zvau's,  clan  on  certain 
special  occasions;  and  this  act,  like  other  honouring  of  the 
wau's,  totems,  will  be  greeted  with  naven  behaviour  upon 
every  occasion  when  it  is  repeated.  Besides  ritual  duties  and 
privileges,  the  laua  will  help  his  wau  in  such  labours  as  house 
building ;  and  in  the  formal  debating  in  the  ceremonial  house, 
the  laua  will  often  speak  on  the  side  of  his  wau's  clan.  All 
such  acts  will  be  hailed  by  the  wau  whenever  they  occur  with 
gestures  and  exclamations  reminiscent  of  naven. 

4.  Boasting  in  the  presence  of  the  wau.  It  is  correct  for  a 
boy  to  boast  in  the  presence  of  his  wau,  but  the  latter  may 
resent  this  behaviour  if  it  is  carried  to  excess;  and  he  will, 
in  this  event,  make  a  gesture  suggestive  of  turning  his  but- 
tocks to  his  laua.  I  have  never  seen  the  complete  gesture  of 
rubbing  the  buttocks  on  the  laud's  shin,  the  climax  of  the 
naven,  carried  out  in  reply  to  laud's  boasts ;  usually  the  threat 
is  sufficient  to  curb  the  young  man's  tongue.  But  it  is 
generally  stated  that,  when  exasperated,  a  wau  may  complete 
the  gesture  and,  by  so  doing,  involve  the  laua  in  a  presentation 
of  valuables  to  the  wau.  It  is  probable  that  this  boasting  in 
the  presence  of  the  wau  should  be  classified  with  the  other 
acts  characteristic  of  the  laua.  But  the  case  diifers  from  the 
others  in  that  the  naven  behaviour  of  the  wau  is  carried  out 
in  anger  or  annoyance.  In  the  other  cases  the  naven  behaviour 
is  on  the  whole  a  method  of  complimenting  or  congratulating 
the  laua  upon  his  achievements. 

5.  Changes  in  social  status.  Such  events  in  the  boy's  life 
as :  the  boring  of  his  ears,  the  boring  of  his  nasal  septum,  his 
initiation,  his  marriage,  his  becoming  possessed  by  a  shamanic 
spirit,  may  all  be  celebrated  with  naven  when  they  occur, 
I  feel  some  hesitation  in  applying  the  term  change  of  status 
to  these  events,  since  the  culture  does  not  clearly  emphasise 


The  Naven  Ceremonies  9 

the  concepts  of  status  and  promotion.  Even  in  the  graded 
initiatory  system,  the  event  which  the  wau  celebrates  is  not 
the  promotion  of  the  boy  from  one  grade  to  the  next,  but  the 
completion  of  the  ceremony  of  initiation  which  may  take 
place  years  before  the  general  shifting  of  age-grade  member- 
ship. In  any  case  it  must  be  clearly  understood  that  the 
naven  ceremonies  are  in  no  sense  rites  de  passage,  although 
they  may  celebrate  or  emphasise  the  fact  that  such  rites  have 
been  performed.  The  actual  rites  of  initiation  are  performed 
by  the  elder  age  grade  within  a  system  of  social  groupings 
and  moieties  which  is  to  a  great  extent  independent  of  clans 
and  totemism.  The  fact  that  the  rite  de  passage  has  been 
performed  is  afterwards  celebrated  by  the  wau  within  the 
different  social  grouping  which  is  organised  on  the  basis  of 
the  family  and  totemic  clans. 

The  other  most  important  events  in  a  man's  life — his  birth 
and  his  death — are  neither  of  them  celebrated  by  naven.  In 
the  case  of  birth,  the  wau  will  go  to  the  child  soon  after  it  is 
born  and  present  it  with  a  coconut  and  a  personal  name  which 
refers  to  the  totemic  ancestors  of  the  wau's  clan.  Though 
I  have  never  seen  this  done,  I  believe  that  the  wau  might 
well  exclaim:  " Lan  men  to!"  (husband  thou  indeed!)  when 
the  baby  gripped  his  fingers.  This  exclamation,  as  we  shall 
see  later,  is  one  of  the  characteristic  details  of  naven  behaviour. 
We  shall  see  too  that  the  giving  of  the  special  name  is  an  act 
which  demonstrates  the  existence  of  the  wau-laua  relation- 
ship ;  and  that  the  coconut  is  the  first  in  a  long  series  of  gift 
exchanges  which  will  accompany  naven  and  other  ceremonies. 
In  the  case  of  death,  there  is  again  no  naven,  but  the  classifi- 
catory  wau  plays  an  important  part  in  the  mortuary  cere- 
monies, and  finally  claims  the  dead  man  as  in  some  special 
sense  a  member  of  the  maternal  clan,  pulling  the  figure  which 
represents  him  towards  himself  with  a  hook.  In  the  land  of 
the  dead  the  ghost  will  henceforth  live  under  the  names  which 
have  been  given  to  him  by  his  wau. 

The  event  of  marriage  may  be  celebrated  not  only  by  wau, 
but  also,  I  believe,  by  tawontu  (wife's  brother).  In  one  of 
the  myths  which  I  collected,  there  occurs  a  casual  mention 
of  the  fact  that  a  bride's  own  brother  rubs  his  buttocks  on 


lo  The  Naven  Ceremonies 

the  bridegroom's  shin.    I  do  not  know  of  any  other  occasion 
which  is  celebrated  in  this  way  by  tawontu. 

There  are  more  naven  and  more  occasions  for  naven  in  the 
life  of  a  boy  or  a  man  than  in  the  Hfe  of  a  girl,  but  the  achieve- 
ments of  a  girl  may  also  be  occasions  for  naven.  The  list  for 
a  girl  includes:  catching  nsh  with  hook  and  line;  collecting 
mayflies ;  washing  sago ;  cooking  sago  pancakes ;  cooking  sago 
paste ;  making  a  fish  trap,  a  rain  cape  or  a  mosquito  bag ;  and 
bearing  a  child.  These  events  all  fall  into  the  category  of 
cultural  acts  which  are  celebrated  at  their  first  accomplish- 
ment. Besides  these  there  are  two  other  events,  initiation  and 
dancing  in  the  tshugukepma  dances,  both  of  which  may  be 
celebrated  with  naven.  The  initiation  of  girls  is  distinct  from 
female  puberty  ceremonies  and  consists  of  scarification  and 
the  showing  of  flutes.  It  is  a  simplified  version  of  the  initia- 
tion of  boys  and  is  carried  out  by  the  men  of  the  elder  age 
grade.  The  ceremony  is  performed  on  only  a  very  few  women. 
The  tshugukepma  ceremonies  are  occasions  on  which  the 
lauas  exhibit  the  totemic  ancestors  of  the  waus  clan  and 
dance  in  masks  which  represent  these  ancestors,  so  that  they 
fall  into  the  category  of  "acts  characteristic  of  laua".  I  do 
not  know  anything  about  girls  boasting  in  the  presence  of 
their  zvaus ;  in  any  case  zvau's  gesture  of  rubbing  his  buttocks 
on  laua's  shin  is  specifically  a  part  of  naven  for  a  boy,  and  an 
analogue  of  this  gesture  in  the  ceremonies  for  a  girl  may  be 
seen  in  the  pantomime  of  her  birth  from  the  belly  of  wau. 


MATERIALS   UPON   WHICH   THE 
DESCRIPTION    IS    BASED 

Although  the  list  of  possible  occasions  is  so  long,  in  actual 
practice  large  naven  ceremonies  are  not  very  often  performed, 
and  their  frequency  is  perhaps  limited  by  the  expense  which 
they  involve.  In  most  instances,  if  the  land's,  achievement  is 
brought  to  the  notice  of  the  zcau,  the  latter  only  exclaims: 
^^ Lan  men  to!"  (husband  thou  indeed!),  throws  some  lime 
over  his  laua  and  hails  him  ceremonially  with  a  string  of 
names  of  ancestors  of  zcaii's  clan.    Besides  these  instances  in 


The  Naven  Ceremonies  ii 

which  the  naveji  ritual  is  reduced  to  mere  gestures  and  ex- 
clamations, there  are  many  naven  which  are  only  celebrated 
on  a  small  scale.  It  is  hard  for  me  to  judge  of  the  frequency 
of  these  small  naven  because  I  may  often  have  heard  nothing 
of  such  occurrences  even  though  they  took  place  while  I  was 
in  the  village.  I  have  only  witnessed  five  nave7i  in  which  any 
of  the  ritual  was  carried  out.  Of  these,  one  was  in  the  village 
of  Mindimbit,  for  a  group  of  children  who  had  been  out 
working  sago  in  the  swamps  and  had  returned,  the  little  boys 
having  accomplished  the  felling  of  the  palm  and  the  beating 
of  the  pith  and  the  little  girls  having  washed  the  resulting 
pulp  to  get  the  sago  (cf.  the  myth  quoted  on  p.  3  6).  I  witnessed 
two  naven  in  Kankanamun  village,  one  of  which  was  for  a 
boy  who  had  made  a  canoe  and  the  other  for  a  young  man 
who  had  killed  a  pig  and  stood  a  feast.  In  both  of  these 
ceremonies  only  the  women  (mothers,  fathers'  sisters,  and 
elder  brothers'  wives)  took  part,  the  waus  apparently  ignoring 
the  occasions.  I  witnessed  one  naven  in  Palimbai  for  a  big 
new  canoe ;  in  this  only  two  waus  took  part  and  the  occasion 
was  ignored  by  the  women.  Lastly,  I  witnessed  an  incident 
in  Malinggai  in  which  a  wau  dashed  into  the  midst  of  a 
tshugukeptna  dance  and  rubbed  his  buttocks  on  the  shin  of 
a  male  laua  who  was  impersonating  one  of  the  wau's  clan 
ancestors.  Besides  these  small  navens — none  of  which  was 
at  all  elaborate — I  obtained  good  descriptions  of  a  larger 
naven  which  took  place  in  Mindimbit  for  a  little  girl  who 
had  caught  a  fish  with  hook  and  line,  and  I  collected,  also  in 
Mindimbit,  good  descriptions  of  the  elaborate  and  varied 
naven  ceremonies  which  used  to  be  held  for  the  successful 
homicide.  My  account  of  naven  is  based  on  this  material 
together  with  scattered  references  to  naven  in  the  conversa- 
tions of  informants  and  in  mythology. 

The  four  villages,  Mindimbit,  Kankanamun,  Palimbai  and 
Malinggai,  are  closely  related  in  language  and  culture,  and 
references  to  the  small  local  differences  in  social  organisation 
and  behaviour  will  be  inserted  when  relevant. 


12  The  Naven  Ceremonies 


DESCRIPTION   OF  THE   CEREMONIES 

The  outstanding  feature  of  the  ceremonies  is  the  dressing 
of  men  in  women's  clothes  and  of  women  in  the  clothes 
of  men.  The  classificatory  wau  dresses  himself  in  the  most 
filthy  of  widow's  weeds,  and  when  so  arrayed  he  is  referred 
to  as  "nyame"  ("mother").  Plates  II  to  IV  show  two 
classificatory  zoaus  arrayed  for  the  naven  of  the  young  man 
in  Palimbai  who  had  made  a  large  canoe  for  the  first  time  in 
his  life.  They  put  on  the  most  filthy  old  tousled  skirts  such 
as  only  the  ugliest  and  most  decrepit  widows  might  wear,  and 
like  widows  they  were  smeared  with  ashes.  Considerable 
ingenuity  went  into  this  costuming,  and  all  of  it  was  directed 
towards  creating  an  effect  of  utter  decrepitude.  On  their 
heads  they  wore  tattered  old  capes  which  were  beginning  to 
unravel  and  to  fall  to  pieces  with  age  and  decay.  Their  bellies 
were  bound  with  string  like  those  of  pregnant  women.  In 
their  noses  they  wore,  suspended  in  place  of  the  little  triangles 
of  mother-of-pearl  shell  which  women  wear  on  festive  occa- 
sions, large  triangular  lumps  of  old  sago  pancakes,  the  stale 
orts  of  a  long  past  meal. 

In  this  disgusting  costume  and  with  absolutely  grave  faces 
(their  gravity  was  noted  with  special  approbation  by  the 
bystanders),  the  two  "mothers"  hobbled  about  the  village 
each  using  as  a  walking  stick  a  short  shafted  paddle  such  as 
women  use.  Indeed,  even  with  this  support,  they  could  hardly 
walk,  so  decrepit  were  they.  The  children  of  the  village 
greeted  these  figures  with  screams  of  laughter  and  thronged 
around  the  two  "mothers",  following  wherever  they  went 
and  bursting  into  new  shrieks  whenever  the  "mothers",  in 
their  feebleness,  stumbled  and  fell  and,  falling,  demonstrated 
their  femaleness  by  assuming  on  the  ground  grotesque  atti- 
tudes with  their  legs  widespread. 

The  "mothers"  wandered  about  the  village  in  this  way 
looking  for  their  "child"  (the  laua)  and  from  time  to  time 
in  high-pitched,  cracked  voices  they  enquired  of  the  by- 
standers to  learn  where  the  young  man  had  gone.  "We  have 
a  fowl  to  give  to  the  young  man"  (cf.  Plate  II).  Actually  the 


The  Naven  Ceremonies  13 

laua  during  this  performance  had  either  left  the  village  or 
hidden  himself.  As  soon  as  he  found  out  that  his  waus  were 
going  to  shame  themselves  in  this  way,  he  went  away  to 
avoid  seeing  the  spectacle  of  their  degraded  behaviour. 

If  the  wau  can  find  the  boy  he  will  further  demean  himself 
by  rubbing  the  cleft  of  his  buttocks  down  the  length  of  his 
laua's,  leg,  a  sort  of  sexual  salute  which  is  said  to  have  the 
effect  of  causing  the  laua  to  make  haste  to  get  valuables  which 
he  may  present  to  his  wau  to  "make  him  all  right ".^  The 
laua  should,  nominally  at  least,  fetch  valuables  according  to 
the  number  of  times  that  the  wau  repeats  the  gesture — one 
shell  for  each  rubbing  of  the  buttocks. 

The  wau's  gesture  is  called  mogul  nggelak-ka.  In  this 
phrase  the  word  mogul  means  "anus",  while  nggelak-ka  is  a 
transitive  verb  which  means  "grooving",  e.g.  ian  nggelak-ka 
means  to  dig  a  ditch.  The  suffix  -ka  is  closely  analogous  to 
the  English  suffix,  -ing,  used  to  form  present  participles  and 
verbal  nouns. 

This  gesture  of  the  wau  I  have  only  seen  once.  This  was 
when  a  wau  dashed  into  the  midst  of  a  dance  and  performed 
the  gesture  upon  his  laua  who  was  celebrating  the  wau's 
ancestors.  The  wau  ran  into  the  crowd,  turned  his  back  on 
the  laua  and  rapidly  lowered  himself — almost  fell — into  a 
squatting  position  in  such  a  way  that  as  his  legs  bent  under 
him  his  buttocks  rubbed  down  the  length  of  the  laua's  leg. 

But  in  the  particular  naven  which  I  am  describing,  the  two 
waus,  did  not  find  their  laua  and  had  to  content  themselves 
with  wandering  around  the  village  in  search  of  him.  Finally 
they  came  to  the  big  canoe  which  he  had  made — the  achieve- 
ment which  they  were  celebrating.  They  then  collapsed  into 
the  canoe  and  for  a  few  moments  lay  in  it  apparently  helpless 
and  exhausted  (Plate  IV),  with  their  legs  wide  apart  in  the 
attitudes  which  the  children  found  so  amusing.  Gradually 
they  recovered  and  picked  up  their  paddles,  and  sitting  in 
the  canoe  in  the  bow  and  stern  (women  sit  to  paddle  a  canoe, 
but  men  stand),  they  slowly  took  it  for  a  short  voyage  on  the 

^  This  is  the  pidgin  English  translation  of  an  latmul  phrase,  kunak-ket. 
The  suffix  -ket  is  purposive ;  and  the  word  kunak  means  to  "  make  ready  ", 
"repair",  or  "propitiate". 


14  The  Naven  Ceremonies 

lake.  When  they  returned  they  came  ashore  and  hobbled  off. 
The  performance  was  over  and  they  went  away  and  washed 
themselves  and  put  on  their  ordinary  garments.  The  fowl 
was  finally  given  to  the  laua  and  it  became  his  duty  to  make 
a  return  present  of  shell  valuables  to  his  ivau  at  some  later 
date.  Return  presents  of  this  kind  are  ceremonially  given, 
(generally  on  occasions  when  some  other  dances  are  being 
performed.  The  shells  are  tied  to  a  spear  and  so  presented 
to  the  wau. 

In  more  elaborate  naven,  especially  those  in  which  women 
play  a  part,  there  is  a  classificatory  spreading  of  ritual  be- 
haviour not  only  to  cause  the  classificatory  relatives  of  the 
actual  laiia  to  perform  naven  for  him,  but  also  to  cause  per- 
sons not  otherwise  involved  to  adopt  7iaven  behaviour  towards 
other  individuals  who  may  be  identified  in  some  way  with 
the  actual  laua.  For  example,  the  characteristic  naven  be- 
haviour of  the  elder  brother's  wives  is  the  beating  of  their 
husband's  younger  brother  when  his  achievements  are  being 
celebrated.  Owing  to  the  classificatory  spreading  of  the  naven 
not  only  does  the  boy  who  has  worked  sago  get  beaten  by 
his  elder  brother's  wives,  but  also  the  boy's  father's  elder 
brothers'  wives  get  up  and  beat  the  father.  Further,  men 
other  than  the  performing  wau?,  may  take  the  opportunity 
to  make  presentations  of  food  to  their  various  lauas,. 

Some  of  this  classificatory  spreading  occurred  in  the  naven 
in  Mindimbit  which  was  celebrated  for  the  children  who 
had  worked  sago  for  the  first  time.  In  this  ceremony  only 
female  relatives  performed  and  their  costume  was  in  sharp 
contrast  to  that  of  the  waus,  described  above.  The  wau  dresses 
himself  in  the  filthiest  of  female  garments,  but  the  majority 
of  the  women,  when  they  put  on  the  garments  of  men,  wear 
the  smartest  of  male  attire.  The  female  relatives  who  per- 
formed were  sisters  (nyanggai),  fathers'  sisters  (lau),  elder 
brothers'  wives  [tshatshi),  mothers  (nyame),  and  mothers' 
brothers'  wives  (mbora) — all  these  terms  being  used  in  a 
classificatory  as  well  as  in  the  narrow  sense. 

Of  these  relatives  the  sisters,  fathers'  sisters  and  elder 
brothers'  wives  dressed  as  men,  borrowing  the  very  best  of 
feather  headdresses   and   homicidal  ornaments   from  their 


The  Naven  Ceremonies  15 

menfolk  (husbands  or  brothers  or  fathers).  Their  faces  were 
painted  white  with  sulphur,  as  is  the  privilege  of  homicides, 
and  in  their  hands  they  carried  the  decorated  lime  boxes  used 
by  men  and  serrated  lime  sticks  with  pendant  tassels  whose 
number  is  a  tally  of  men  killed  by  the  owner.  This  costume 
was  very  becoming  to  the  women  and  was  admired  by  the 
men.  In  it  the  women  were  very  proud  of  themselves.  They 
walked  about  flaunting  their  feathers  and  grating  their  lime 
sticks  in  the  boxes,  producing  the  loud  sound  which  men  use 
to  express  anger,  pride  and  assertiveness.  Indeed  so  great 
was  their  pleasure  in  this  particular  detail  of  male  behaviour 
that  the  husband  of  one  of  them,  when  I  met  him  on  the  day 
following  the  performance,  complained  sorrowfully  that  his 
wife  had  worn  away  all  the  serrations  on  his  lime  stick  so 
that  it  would  no  longer  make  a  sound. 

The  mothers  and  mothers'  brothers'  wives  wore  different 
types  of  costume.  The  mother  stripped  off  her  skirt  but  did 
not  put  on  any  male  ornaments,  and  the  mothers'  brothers' 
wives  put  on  filthy  widow's  weeds,  like  those  of  the  zcauy 
described  above. 

With  these  modifications  of  costume  there  goes  also  a 
modification  of  the  kinship  terms  used  for  these  women. 
Thus  : 

iau  (father's  sister)  becomes  ''iau-ndo  "  or  "nyai' "  (literally 

"father"). 
tshaishi  (elder   brother's   wife)   becomes    tshaishi-ndo   or 

nyatnun  (literally  "elder  brother"). 
mbora  (mother's  brother's  wife)  becomes  mbora-ndo. 
nyanggai  (sister)  becomes  nyanggai-ndo. 

In  these  special  terms  the  suffix  -ndo  means  "man".  The 
mothers,  who  alone  of  the  female  relatives  concerned  in  naveti 
put  on  no  male  attire,  are  still  described  as  nyame  (mother). 
The  children's  canoe  came  back  from  the  sago  swamps  late 
in  the  morning  and  as  soon  as  it  was  sighted  from  the  banks 
of  the  river,  word  was  shouted  in  to  the  village  which  lies 
back  from  the  river  on  a  little  lake.  The  women  assembled 
on  the  shores  of  the  lake  and  when  the  canoe  entered,  they 
swam  out  to  greet  and  splash  the  children,  as  is  done  when 


1 6  The  Naven  Ceremonies 

a  canoe  returns  from  a  successful  head-hunting  raid.  When 
the  children  had  landed,  the  village  appeared  to  go  mad  for 
awhile;  fathers'  sisters  and  elder  brothers'  wives  dashing 
about  searching  respectively  for  their  various  brothers'  chil- 
dren and  husbands'  younger  brothers  in  order  to  beat  them. 
The  men  who  were  expecting  to  be  beaten  did  their  best  to 
avoid  the  ceremony  by  skulking  in  the  ceremonial  houses, 
but  the  women  on  these  occasions  have  unusual  licence  (per- 
haps because  they  are  in  men's  costume).  Whenever  a  iau 
saw  her  kanggat  (brother's  child,  w.s.)  or  a  tshaishi  saw  her 
tshuambo  (husband's  younger  brother)  in  one  of  the  cere- 
monial houses,  she  dashed  into  the  usually  forbidden  spot, 
stick  in  hand,  and  gave  him  several  good  blows;  and,  if  he 
ran  away,  she  chased  him,  beating  him  as  he  ran.  On  a 
similar  occasion  in  Kankanamun,  when  the  women  were 
celebrating  naven  for  a  young  man  who  had  killed  a  pig  and 
was  standing  a  feast,  they  hesitated  to  enter  a  ceremonial 
house  in  which  a  debate  was  in  progress.  The  men  stopped 
the  debate  for  their  benefit  and  the  whole  crowd  of  them 
danced  into  the  ceremonial  house  in  a  column.  Arrived  there, 
the  colurnn  broke  up  and  each  woman  went  and  beat  her 
appropriate  relatives.^ 

On  the  occasion  in  Mindimbit  which  I  am  describing,  I  did 
not  see  any  activity  of  the  mothers  and  sisters.  The  activity 
of  the  other  women  dressed  as  men  continued  off  and  on 
throughout  the  rest  of  the  day.  In  the  evening,  the  women 
had  a  small  dance  by  themselves.  It  is  the  custom  of  the 
men  to  take  off  their  pubic  aprons  after  dark  and  the  women 
accordingly  stripped  off  the  flying-fox-skin  aprons,  which 
they  had  borrowed  from  their  husbands  and  brothers,  and 
danced  with  their  loins  uncovered,  still  wearing  their  splendid 
feather  headdresses  and  ornaments.  There  was  no  apparent 
embarrassment  of  the  men  at  this  exposure  in  their  women- 

^  To  this  account  of  the  beatings,  two  points  may  be  added  which 
I  did  not  observe  at  the  time  but  of  which  I  was  later  informed.  It  is 
said  that  the  actual  father  of  the  hero  would  only  be  slightly  beaten  by 
his  tshaishi  but  that  the  father's  younger  brother  (tshambwi-nyai')  would 
be  severely  beaten.  I  was  also  told  that  on  these  occasions  the  tshaishi 
will  exclaim  "  tshtiambo-ket  wonggegio" — (I  will  rape  my  young 
brother). 


The  Naven  Ceremonies  ly 

folk — but  the  older  men  were  rather  shocked  at  this  lack  of 
embarrassment.  The  dance  was  held  close  to  one  of  the 
smaller  ceremonial  houses  and  one  of  the  old  men  remarked 
to  me  in  scandalised  tones  that  it  was  "shocking"  to  see  the 
younger  men  crowding  to  this  ceremonial  house,  where  they 
had  no  business  but  a  good  view  of  the  women's  dance. 
It  was  the  ceremonial  house  of  a  particular  clan,  but  that 
evening  it  was  full  of  young  men,  members  of  every  clan  in 
the  village. 

In  the  case  of  the  naven  in  Mindimbit  for  the  little  girl 
who  had  caught  a  fish  with  hook  and  line,  the  performance  was 
still  more  elaborate,  not  because  the  achievement  was  very 
important  but  perhaps  because  the  waus  were  anxious  to 
obtain  shell  valuables  by  presenting  pigs  to  their  lauas.  From 
the  accounts  which  I  collected,  it  appears  that  in  this  naveti 
both  male  and  female  relatives  took  part  and  eight  pigs  were 
killed.  Besides  the  little  girl  who  had  caught  a  fish  (a  month 
or  two  before  the  ceremony),  two  other  little  girls  were 
honoured,  and  the  resulting  naven  spread  to  almost  every 
individual  in  the  village.  One  pig  was  even  presented  to  a 
classificatory  laua  in  the  next  village. 

Four  zoaus  were  dressed  as  "  mothers  "  and  their  skirts  were 
tucked  up  so  as  to  expose  their  genitals.  Three  of  these  waus 
are  described  as  carrying  the  little  girls  "on  their  heads" — 
presumably  carrying  girls  in  the  position  in  which  mothers 
habitually  carry  their  children  astride  on  the  shoulder  (cf. 
Plate  V  b).  The  fourth  zvau  was  also  dressed  as  a  mother  but 
wore  no  skirt.  He  was  tied  down  on  some  sort  of  "bed"  or 
stretcher,  on  which  he  was  lifted  and  violently  swung  by  a 
number  of  men  who,  while  they  rocked  the  stretcher,  sang 
songs  of  the  wau's  clan.  The  little  girl  who  had  caught  the 
fish  was  placed  on  the  belly  of  this  wau,  her  father  meanwhile 
standing  by  with  an  adze  to  which  he  had  tied  a  mother-of- 
pearl  crescent.  This  adze  he  gave  into  the  little  girl's  hands 
and  with  it  she  cut  the  bonds  which  held  her  wau  to  the 
stretcher.  She  then  gave  the  decorated  adze  to  the  wau  and 
with  it  he  raised  himself  to  a  sitting  position  on  the  stretcher, 
supporting  himself  by  means  of  the  adze.  Similarly  the 
fathers  of  the  other  little  girls  presented  valuables  to  the 


1 8  The  Naven  Ceremonies 

girls'  waus  and  at  the  same  time  loosed  the  waus'  skirts.  The 
waus  then  resumed  their  normal  male  attire. 

During  the  above  performance  a  mbora,  the  wife  of  the 
wau  who  was  tied  to  the  stretcher,  danced  with  her  skirts 
tucked  up  to  expose  her  genitals.  She  wore  a  string  bag  over 
her  head  and  face,  and  carried  a  digging  stick,  holding  it 
horizontally  behind  her  shoulders  with  her  hands  raised  to 
hold  it  on  each  side.  At  the  conclusion  of  the  dance  she  was 
presented  with  a  mother-of-pearl  crescent  and  three  Turbo 
shells  by  the  father  of  the  little  girl. 

The  whole  of  this  ritual  pantomime  acted  by  the  wau 
appears  to  me  to  be  a  representation  of  the  birth  of  the  little 
girl  from  the  belly  of  her  mother's  brother,  though  none  of 
my  male  informants  (in  Kankanamun  where  I  made  enquiries) 
had  ever  heard  of  the  custom  of  tying  the  woman  in  labour 
on  a  stretcher.  The  dance  of  the  mbora  (mother's  brother's 
wife),  with  her  arms  stretched  back  behind  her  head,  also 
probably  represents  the  position  of  a  woman  in  labour. 

A  general  presentation  of  food  and  valuables  followed  this 
ceremony.  Eight  pigs  were  killed  and  presented.  Of  these 
one  was  given  by  her  wau  to  the  little  girl  who  had  caught  a 
fish,  and  she  ceremonially  stepped  upon  it.  Of  the  remaining 
pigs,  three  were  given  by  wau?,  to  classificatory  lauas,  in- 
cluding one  pig  which  was  given  to  the  donor's  sister's  hus- 
band's father,^  in  the  next  village.  One  of  the  other  pigs  was 
given  by  a  woman  to  her  husband's  sister's  child  (i.e.  mbora 
gave  to  nasa,  husband's  sister's  child).  One  was  given  to  a 
classificatory  sister's  husband  (i.e.  to  the  potential  father  of  a 
laua),  and  one  was  given  to  a  widowed  classificatory  sister.^ 

^  This  relative  also  is  called  laua.  The  latmul  kinship  system  is  charac- 
terised by  an  alternation  of  generations  similar  to  that  found  in  many 
Australian  systems.  Theoretically  a  man  is  re-incarnated  in  his  son's  son; 
and  a  woman  in  her  brother's  son's  daughter.  In  general  the  kinship 
term  which  is  appropriate  for  a  given  relative  will  also  be  appropriate 
for  any  other  relative  of  the  same  sex  in  an  analogous  position  two 
patrilineal  generations  away.  Thus  the  term  nyarnun  is  used  both  for  elder 
brother  and  for  father's  father.  The  term  nyame  is  used  both  for  mother 
and  for  son's  wife.  In  the  present  instance  the  term  laua  is  used  both 
for  sister's  husband's  son  and  sister's  husband's  father. 

^  According  to  the  collected  account,  the  eighth  pig  was  given  by  a 
man  to  his  own  mother's  brother's  son.  This  is  anomalous  and  probably 
either  the  account  is  at  fault  or  the  two  persons  concerned  were  also 
related  in  an  inverse  way  through  some  other  route. 


The  Naven  Ceremonies  19 

We  have  seen  above  that  the  father  of  the  laua  provided 
the  valuables  which  the  little  girl  gave  to  her  wau.  In  the 
general  presentation  of  food  and  valuables  this  pattern  is 
further  extended  so  that  the  exchanges  actually  take  place 
between  the  relatives  of  wives  on  the  one  hand  and  the  rela- 
tives of  husbands  on  the  other.  The  wife's  clan  and  her 
classificatory  brothers,  etc.,  give  pigs  to  the  husband's  son  or 
to  the  husband's  father  or  to  the  husband.  This  system  is 
summed  up  in  the  native  language  by  saying  that  the  wau- 
nyame  nampa  (mother's  brother  and  mother  people)  give  pigs 
to  the  lanoa  nampa  (husband  people)  or  laua  nyanggu  (sisters' 
children  people) — these  two  terms  being  almost  synonymous 
(cf.  p.  93).  In  the  return  presentation  of  valuables,  the 
classificatory  brothers  and  other  relatives  of  the  recipients 
of  pigs  contributed  to  the  accumulation  of  valuables  which 
made  up  the  return  gifts,  so  that  the  list  was  exceedingly 
elaborate  and  involved  a  great  number  of  people,  related  in 
the  most  various  ways  to  the  original  donors  of  the  pigs. 

The  last  naven  to  which  I  shall  refer  is  that  which  used  to 
be  celebrated  for  the  successful  homicide  when  he  had  accom- 
plished his  first  kill.  The  account  which  I  have  of  these  cere- 
monies repeats  some  of  the  material  which  I  have  already 
given,  but  it  also  describes  four  other  pantomime  incidents 
which  may  be  celebrated  on  such  occasions.  The  incidents 
are  apparently  independent  one  of  the  other,  but  all  of  them 
involve  the  same  types  of  transvesticism  which  I  have  de- 
scribed above.  The  only  exception  to  this  is  the  costume  of 
mbora,  who  though  she  wears  ragged  female  attire  in  the 
first  incident  puts  on  (probably  ragged^)  male  attire  in  the 
second  and  third. 

In  the  first  incident,  the  mbora  dances  in  bedraggled  skirt 
with  her  head  inside  a  fishing  net.  She  wears  the  enemy 
head  suspended  from  her  neck,  and  with  her  raised  hands  she 
holds  a  digging  stick  behind  her  shoulders.  Another  mbora 
might  have  the  jaw  bone  from  the  enemy  head  hung  around 
her  neck.  This  dance  continues  until  sunset  when  the  nasa 

^  Unfortunately  I  do  not  know  whether  the  mbora  wears  smart  or 
bedraggled  costume  when  dressed  as  a  male — but  I  suspect  that  she,  like 
her  husband,  wears  filth. 


20  The  Naven  Ceremonies 

(husband's  sister's  son)  brings  a  Turbo  shell  tied  to  a  spear 
and,  presenting  this,  looses  the  old  skirt  from  his  mhora. 

In  the  second  incident,  the  iau-ndo  (father's  sisters  in  male 
costume)  carry  in  their  hands  a  feather  ornament  such  as 
the  homicide  will  wear  in  his  headdress.  The  mhora-ndo 
(mother's  brother's  wives  in  male  costume)  lie  down  on  the 
ground  and  the  iau-ndo  go  stepping  across  them  carrying  the 
feathers.  The  mhora-ndo  then  snatch  the  feathers  and  run 
off  with  them. 

In  the  third  incident,  the  zvau  puts  on  a  skirt  and  fixes 
an  orange-coloured  fruit,  mbuandi  {Ervatamia  aurantiaca),  in 
his  anus  and  goes  up  the  ladder  of  a  house  displaying  this  as 
he  climbs.  At  the  top  of  the  ladder  he  goes  through  the 
actions  of  copulation  with  his  wife,  who  is  dressed  and  acts 
as  the  male.  The  laua  is  much  ashamed  at  this  spectacle  and 
the  laua's  sister  would  weep  at  it.  The  orange  fruit  represents 
an  anal  clitoris,  an  anatomical  feature  frequently  imagined 
by  the  latmul  and  appropriate  to  the  waus  assumption  of 
grotesque  feminity.  My  informant  told  me  that,  after  the 
mhora,  acting  as  male,  had  copulated  with  the  wan,  the  other 
women  would  all  follow  suit — and  we  may  imagine  a  scene 
of  considerable  confusion  around  the  unfortunate  wau. 

In  the  fourth  incident,  a  large  pear-shaped  prawn  trap 
wuth  lobster  pot  entrance  is  put  on  the  ladder  of  a  house. 
All  the  women  of  the  village  lie  down  naked  on  the  ground, 
side  by  side  in  front  of  the  ladder.  The  killer  steps  over  the 
women  to  go  up  into  the  house.  As  he  walks,  he  is  ashamed 
to  look  at  their  genitals ;  he  therefore  goes  with  his  head  up, 
feeling  his  way  with  his  feet.  On  this  occasion  the  women 
would  say  of  the  vulva:  "  That  so  small  a  place  out  of  which 
this  big  man  came."  The  only  women  over  whom  he  would 
not  step  are  his  sister  and  his  wife.  His  sister  would  accom- 
pany him  as  he  steps.  She  shows  no  modesty  but  attacks 
the  vulvae  with  her  hands  as  she  passes — especially  that  of 
tshaishi,  the  elder  brother's  wife.  Seeing  this,  she  would 
exclaim:  "A  vulva!"  But  tshaishi  would  reply:  "No!  A 
penis!" 

After  this  the  tshaishi  sings  a  comic  song,  beating  time 
with  a  coconut  shell.  The  text  of  this  song,  like  that  of  all 


The  Naven  Ceremonies  21 

latmul  songs,  is  built  upon  a  series  of  totemic  names,  in  this 
case,  names  of  the  fish  trap  set  on  the  ladder. 

This  fish  trap,  Alie-namak^  fish  trap. 

Go !  you  will  never  shoot  it. 

Go!  you  will  never  spear  it. 

This  is  Woli's^  fish  trap. 

This  is  Tanmbwa's^  fish  trap. 

Go !   You  will  never  shoot  it. 

Go!   You  will  never  spear  it.   Etc. 

This  is  repeated  substituting  other  names  for  Alie-namak  in 
the  first  line.  At  the  close  of  the  song  the  killer  spears  the 
trap  and  goes  up  into  the  house. 

In  this  last  incident,  the  fish  trap  is  certainly  a  symbol  of 
the  vulva,  and  this  symbol  occurs  also  in  esoteric  mythology, 
together  with  the  complementary  symbol,  the  eel,  which  gets 
caught  in  the  trap. 

Before  attempting  to  account  for  the  details  of  the  cere- 
monial, we  may  conveniently  summarise  the  naveti  behaviour 
of  the  various  relatives : 

Wau  (mother's  brother)  wears  grotesque  female  attire; 
off"ers  his  buttocks  to  male  laua;  in  pantomime  gives  birth 
to  female  laua  who  looses  his  bonds;  supports  himself  on 
the  adze  presented  by  her;  presents  food  to  laua  of  either 
sex  and  receives  in  return  shell  valuables;  acts  as  female  in 
grotesque  copulation  with  mbora.  These  ceremonial  acts  may 
be  performed  either  by  own  wau  or  classificatory  wau — most 
usually  the  latter. 

Mbora  (mother's  brother's  wife)  wears  either  grotesque 
female  attire  or  (probably  grotesque)  male  attire ;  dances  with 
digging  stick  behind  her  head;  takes  male  part  in  mimic 
copulation  with  wau ;  like  wau  she  presents  food  to  the  hero 
of  the  naven  and  receives  valuables  in  return. 

lau  (father's  sister)  wears  splendid  male  attire;  beats  the 
boy  for  whom  naven  is  celebrated ;  steps  across  his  prostrate 
mbora ;  participates  in  mimic  contest  between  mbora  and  iau 

^  Alie-namak:  this  is  a  totemic  name  for  a  fish  trap,  an  ancestor  of 
Mandali  clan.  The  meaning  of  the  word,  alie,  is  unknown  to  me.  The  word, 
namak,  is  an  old  term  for  naniivi — fish  trap. 

*■  '  Woli  and  Tanmbwa:  these  are  two  names  for  the  East  Wind  and 
are  used  in  the  jargon  of  shamans  to  refer  to  women.  Each  of  these  lines 
therefore  means:  "This  is  a  woman's  fish  trap." 


22  The  Naven  Ceremonies 

in  which  the  former  snatch  the  feather  headdress  from  the 
latter. 

Nyame  (mother)  removes  her  skirt  but  is  not  transvestite ; 
Hes  prostrate  with  other  women  when  the  homicide  steps 
across  them  all. 

Nyanggai  (sister)  wears  splendid  male  attire ;  accompanies 
the  homicide,  her  brother,  when  he  steps  across  the  women ; 
he  is  ashamed  but  she  attacks  their  genitals,  especially  that 
of  tshatshi;  weeps  when  wau  displays  anal  clitoris. 

Tshaishi  (elder  brother's  wife)  wears  splendid  male  attire ; 
beats  husband's  younger  brother  and  her  vulva  is  attacked 
by  his  sister. 

Tawontu  (wife's  brother,  i.e.  wan  of  ego's  children)  pre- 
sents food  to  sister's  husband  (lando)  and  receives  valuables 
in  return.  Tawontu  may,  I  believe,  rub  his  buttocks  on  lando's 
shin,  when  the  latter  marries  tawontu  s  own  sister. 


CHAPTER  III 

The  Concepts  of  Structure  and  Function 

STRUCTURE 

I  HAVE  described  the  naven  ceremonies  with  a  minimum 
of  references  to  the  culture  in  which  they  are  set;  and 
any  attempt  to  explain  them  must  take  the  form  of  relating 
the  ceremonies  to  their  setting.  Even  to  construct  an  historical 
theory  of  their  origin  would  require  an  exhaustive  examina- 
tion of  the  remainder  of  the  culture,  and  this  is  still  more 
true  of  any  theorising  which  would  account  for  the  naven  in 
terms  of  the  structure  or  of  the  functioning  mechanisms  of 
the  culture.  Since  I  shall  try  to  give  analyses  of  the  position 
of  naven  from  both  structural  and  functional  points  of  view, 
it  will  be  well  to  state  definitely  what  I  mean  at  this  stage^ 
by  the  two  concepts,  Structure  and  Function. 

The  anthropologist  in  the  field  collects  details  of  culturally 
standardised  behaviour.  A  large  part  of  this  material  takes 
the  form  of  native  statements  about  behaviour.  Such  state- 
ments may  be  seen  as  themselves  details  of  behaviour ;  or,  more 
cautiously,  we  may  regard  them  as  true  and  supplementary 
to  the  anthropologist's  account  of  behaviour  which  he  has 
witnessed. 

We  can  view  any  such  detail  from  a  structural  point  of 

^  This  chapter  contains  a  statement  of  my  theoretical  position  when 
I  was  writing  the  body  of  this  book.  The  various  descriptive  chapters  were 
a  series  of  experiments  in  the  use  of  the  different  abstractions  which  are 
here  discriminated  and  defined.  On  the  whole  these  abstractions  have 
stood  the  test  of  experiment,  and  their  use  has  led  to  generalisations  which 
I  have  thought  interesting. 

The  meaning  which  I  attach  to  the  terms  affective  and  cognitive  is  here 
left  vague,  and  the  whole  book  was  written  with  a  vague  use  of  these  terms. 
I  hoped  that  experiments  in  the  description  of  culture  might  clear  up  this 
matter  which  the  study  of  individual  psychology  has  left  obscure.  In 
the  Epilogue  from  a  comparison  of  my  various  methods  I  have  gone  on 
to  attempt  the  solution  of  this  problem,  and  the  position  at  which  I  finally 
arrive  is  slightly  different  from  that  which  is  outlined  in  this  chapter. 
To  re-write  my  statement  of  working  hypotheses  in  terms  of  the  results 
of  my  experiments  would  take  away  from  the  "detective  interest"  of  the 
theoretical  chapters. 


24         The  Concepts  of  Structure  and  Function 

view.  Thus  when  a  mother  gives  food  to  her  child,  we  can 
see  impHcit  in  this  cultural  act  a  number  of  structural  assump- 
tions: that  mothers  feed  children;  that  children  are  depen- 
dent on  their  mothers;  that  mothers  are  kind;  that  taro  is 
edible;  etc.  etc.  When  we  study  a  culture,  we  find  at  once 
that  the  same  structural  assumptions  are  present  in  large 
numbers  of  its  details.  Thus  in  the  behaviour  of  all  latmul 
mothers  we  can  collect  a  series  of  diverse  details,  all  of  which 
imply  that  women  look  after  children;  and  in  a  structural 
enquiry  we  may  regard  the  phrase  "look  after"  as  a  sort  of 
shorthand  by  which  we  may  refer  to  one  structural  aspect 
of  all  these  details  of  cultural  behaviour.  Such  shorthand 
is  not  the  creation  solely  of  the  anthropologist ;  every  culture 
contains  generalised  concepts  which  are  a  shorthand  means 
of  referrmg  collectively  to  structural  aspects  of  large  numbers 
of  details  of  standardised  behaviour.  In  this  structural  ap- 
proach, we  shall  see  such  phrases  as  "mother",  "moiety", 
"patriliny",  etc.,  as  generalised  abstractions  for  referring  to 
the  structural  aspects  of  large  masses  of  cultural  details,  i.e. 
details  of  standardised  behaviour. 

Used  in  this  sense,  the  term.  Structure,  is  closely  related 
in  meaning  to  the  term  "tradition".  But  the  latter  term  has 
disadvantages ;  it  is  evidently  out  of  place  in  any  synchronic 
analysis,  since  in  such  an  analysis  the  historical  origin  of  any 
cultural  detail  is  irrelevant.  In  spite  of  this  the  word  "tradi- 
tion "  does  sometimes  creep  into  synchronic  studies.  In  such 
contexts  the  word  must  clearly  be  stripped  of  all  reference 
to  the  past  or  to  diachronic  processes.  When  so  stripped, 
"tradition"  can  only  mean  t\vQ  given  facts  of  a  culture,  facts 
which  are  "given"  as  premises.  Thus  it  comes  about  that 
the  word  "tradition"  is  sometimes  used  in  synchronic 
analysis  as  a  synonym  for  "  cultural  structure  "  as  here  defined. 

The  word,  structure,  is  a  convenient  collective  term  for 
this  aspect  of  a  culture,  but  it  has  the  disadvantage  that  it 
cannot  conveniently  be  used  to  refer  severally  to  the  elements 
of  which  structure  is  built.  I  shall  therefore  use  the  word 
premise  for  these  elements.  Thus  a  premise  is  a  generalised 
statement  of  a  particular  assumption  or  implication  recognisable 
in  a  number  of  details  of  cultural  behaviour. 


The  Concepts  of  Structure  and  Function         25 

We  shall  see  that  there  is  a  great  deal  of  variation  in  the 
extent  to  which  the  premises  of  a  culture  are  explicitly  stated 
by  the  people;  that  in  many  cases  they  may  only  be  stated 
in  symbolic  terms — by  the  terminology  of  kinship,  by  meta- 
phoric  cliches,  more  obscurely  by  ritual,  or,  in  latmul  cul- 
ture, by  some  trick  in  the  giving  of  personal  names.  But  in 
spite  of  this  variation,  it  is  convenient  to  use  the  term  for 
any  general  statement  whose  expression  can  be  directly  col- 
lected from  informants  or  which  can  be  shown  over  and  over 
again  to  be  implicit  in  the  behaviour  of  individuals. 

The  existence  of  premises  or  "formulations"  in  culture 
has,  of  course,  always  been  recognised,  but  it  is  not  sufficiently 
realised  to  what  extent  the  various  premises  of  a  culture  are 
built  together  into  a  coherent  scheme.  The  student  of  cul- 
tural structure  can  detect  syllogisms  of  the  general  type: 
"A  mother  gives  food  to  children,  a  mother's  brother  is 
identified  with  the  mother,  therefore  mother's  brothers  give 
food  to  children."  And  by  these  syllogisms  the  whole  struc- 
ture of  a  culture  is  netted  together  into  a  coherent  whole. 
As  I  see  it,  the  task  of  structural  anthropology  is  the  investiga- 
tion of  these  schemes  in  culture,  and  I  would  define  cultural 
structure  as  a  collective  term  for  the  coherent  "logical"^  scheme 
which  may  be  constructed  by  the  scientist^  fitting  together  the 
various  premises  of  the  culture. 

Besides  this  use  of  the  word,  structure,  to  refer  to  an  aspect 
of  culture,  there  is  a  second  perfectly  valid  use  of  the  term 
to  refer,  not  to  culture,  but  to  society.  It  is  in  this  sense  that 
the  word  is  chiefly  used  by  Radcliff"e- Brown  and  it  is  im- 
portant to  keep  the  two  senses  distinct.  In  the  study  of 
cultural  structure  we  take  details  of  behaviour  as  our  units 
and  see  them  as  linked  together  into  a  "logical"  scheme; 
whereas  in  the  study  of  social  structure  we  shall  take  human 

^  In  this  definition  I  certainly  am  not  using  the  term  "logic"  in  that 
strict  sense  in  which  it  is  used  in  our  own  culture  to  describe  a  discipline 
of  thought  in  which  the  species  of  steps  involved  are  rigidly  and  con- 
sciously controlled.  I  do  mean,  however,  that  the  elements  of  structiire 
are  linked  together  by  steps.  But  it  is  probable  that  cultures  may  vary 
in  the  species  of  steps  which  link  their  premises  together,  and  that  the 
word  "logic"  must  therefore  be  interpreted  differently  In  every  culture. 
This  concept,  the  eidos,  of  a  culture  will  be  elaborated  at  a  later  stage 
in  this  work. 


26         The  Concepts  of  Structure  and  Function 

individuals  as  our  units  and  see  them  as  linked  together  into 
groups — e.g.  as  kin,  as  clan  members  or  as  members  of  a 
community.  Such  a  difference  as  that  drawn  (p.  97)  be- 
tween peripheral  and  centripetal  orientation  of  social  systems 
might,  for  example,  be  described  as  a  difference  in  social 
structure. 

Thus  in  the  two  disciplines,  the  sociological  and  the  cul- 
tural, we  shall  to  a  great  extent  be  studying  the  same  phe- 
nomena, but  looking  at  them  from  two  different  points  of 
view.  In  the  study  of  cultural  structure  we  shall  see  clans 
and  kinship  terminology  as  shorthand  references  to  details 
of  behaviour,  while  in  the  study  of  social  structure  we  shall 
see  these  groupings  as  segments  in  the  anatomy  of  the  com- 
munity, as  a  part  of  the  mechanism  by  which  the  community 
is  integrated  and  organised. 

For  the  present  I  do  not  see  my  way  to  a  clear  isolation  of 
the  study  of  social  function  from  that  of  social  structure,  and 
I  believe  that  the  most  useful  approach  to  these  disciplines 
which  we  have  at  our  disposal  is  that  which  sees  social 
structure  in  relation  to  the  integration  and  disintegration 
of  the  society.  Therefore  I  shall  consider  these  two  subjects 
together  under  the  general  heading  "Sociology  of  naven". 
But  it  is  easy  and,  I  believe,  valuable  to  isolate  the  study  of 
cultural  structure  from  that  of  other  aspects  of  the  culture 
and  these  subjects  I  shall  therefore  treat  separately. 

FUNCTION 

The  term,  Function,  is  more  difficult  to  define  since  it  is 
ambiguous.  On  the  one  hand  we  have  the  wide  philosophical 
use  of  the  term  to  cover  the  whole  play  of  synchronic  cause 
and  effect  within  the  culture,  irrespective  of  any  consideration 
of  purpose  or  adaptation.  Using  the  term  in  this  sense  we 
might  say  that  the  naven  ceremonies  are  a  function  of  all 
those  elements  of  culture  or  properties  of  society  upon  whose 
presence  the  existence  of  the  ceremonies  depends.  If,  further, 
it  can  be  shown  that  the  naven  ceremonies  play  some  part  in 
the  integration  of  the  community,  we  may  say  that  the  size 
of  the  latmul  villages  is  a  function  of — amongst  other  things — 


The  Concepts  of  Structure  and  Function         27 

the  naven  ceremonies.  This  is  a  perfectly  consistent  and  logical 
use  of  the  term  and  one  which  anthropologists  might  do  well 
to  adopt.i  They  might  then  insert  qualifying  adjectives  to 
indicate  what  sort  of  interdependence  they  are  discussing.  We 
should  then  have  such  phrases  as :  structural  function,  social 
function,  pragmatic  function,  etc. 

But  unfortunately  the  word,  function,  may  also  be  used  in 
the  popular  sense  to  mean  useful  adaptive  effect  and  it  is  often 
not  perfectly  clear  in  which  of  these  senses  the  word  is  used 
in  modern  anthropology.  Malinowski  tends  to  define  func- 
tion in  terms  of  adaptation,  to  regard  all  the  elements  of 
culture  as  "working  directly  or  indirectly  for  the  satisfaction 
of  human  needs  ",  and  he  deduces  from  this  that  every  detail 
of  culture  is  "at  work,  functioning,  active,  efficient ".^  This 
method  of  approach  is  probably  sound  and  its  careful  in- 
vestigation might  give  a  coherent  system  of  anthropology 
allied  to  systems  of  economics  based  upon  "  calculating  man  ". 
But  unfortunately  the  other  sense  of  the  word,  function,  has 
confused  the  issue.  Thus  we  find:  "The  simplest  as  well  as 
the  most  elaborate  artifact  is  defined  by  its  function,  the 
part  that  it  plays  within  a  system  of  human  activities;  it  is 
defined  by  the  ideas  that  are  connected  with  it  and  by  the 
values  which  surround  it."^  From  this  we  must  suppose  that 
"function"  is  here  being  used  not  only  in  reference  to  the 
satisfaction  of  human  needs  but  also  in  reference  to  the  inter- 
dependence of  elements  of  culture. 

The  position  is  complicated  and  difficult;  and  for  me  at 
least  it  is  not  simplified  by  the  theory  that  culture  may  be 
subdivided  into  "institutions",  where  each  institution  is 
defined  by  its  special  function  or  functions.  To  me  it  seems 
premature  to  attempt  such  an  analysis  into  institutions  until 
the  concepts  covered  by  the  term,  function,  have  been 
analysed, 

^  The  chief  difficulty,  which  this  use  involves,  is  that  on  this  principle 
the  naven  ceremonies  could  be  shown  to  be  a  function  of  everything  else 
in  the  universe  and  the  anthropologist  will  be  faced  with  the  problem 
of  defining  his  spheres  of  relevance.  The  attempt  to  subdivide  culture 
into  institutions  is,  no  doubt,  designed  to  delimit  spheres  of  relevance. 

2  Both  quotations  are  from  the  Enc.  Soc.  Set.  1931,  art.  Culture, 
p.  625. 

*  Loc.  cit.  p.  626. 


28         The  Concepts  of  Structure  and  Function 

But  if,  instead  of  emphasising  the  subdivision  of  culture, 
we  examine  the  possibilities  of  subdividing  and  classifying 
different  types  of  functions,  I  believe  that  some  degree  of 
clarity  can  be  achieved  and  that  we  shall  then  see  some  chance 
of  unravelling  the  laws  of  social  and  cultural  functioning. 

A  short  and  incomplete  list  of  the  chief  kinds  of  effect  to 
which  the  term,  function,  has  been  applied  will  suffice  to 
demonstrate  the  present  confusion  and  the  sort  of  modifica- 
tion which  I  am  suggesting : 

1 .  The  direct  satisfaction  of  human  needs. 

2.  The  indirect  satisfaction  of  human  needs. 

3.  The  modification,  elaboration,  etc.,  of  human  needs. 

4.  The  moulding  and  training  of  human  beings. 

5.  The  integration  of  groups  of  human  beings. 

6.  Various  sorts  of  interdependence  and  relationship  be- 
tween elements  of  culture. 

7.  The  maintenance  of  the  status  quo.   Etc.^ 

In  such  a  list  the  fact  which  chiefly  strikes  us  is  that  every  one 
of  these  types  of  function  is  dependent  upon  every  other  type. 
It  is  this  fact  which  has  so  impressed  the  functionalists  that 
they  have  fought  shy  of  analysing  the  concept  of  function. 

Scientific  anthropology  was  born  in  an  age  when  the  older 
sciences  had  so  far  solved  their  domestic  problems  that  they 
were  beginning  to  extend  their  enquiries  into  the  borderline 
areas  which  separate  one  science  from  another.  The  result 
of  this  extension  has  been  a  new  sense  of  the  fundamental 
unity  of  science  and  of  the  world.  But  the  effects  upon  anthro- 
pology have  been  disastrous.  The  emphasis  upon  unity  has 
retarded  analysis.  Physics  and  chemistry  are  benefiting  greatly 
from  a  pooling  of  their  problems  and  methods,  but  this  does 
not  mean  that  they  did  not  benefit  by  their  separation  in  the 
past.  The  great  advances  in  knowledge  are  made  by  analysis 
of  the  problems,  by  the  separation  of  one  class  of  problem 
from  another.  To  state  that  all  problems  are  inter-related  is 
mystical  and  unprofitable.    To  isolate  them  is  artificial  but 

^  The  excellence  of  the  field  work  done  by  members  of  the  functional 
school  is  no  doubt  due  to  the  width  of  the  net  which  they  throw  into  the 
sea  of  culture.  I  am  not  here  advocating  any  reduction  of  this  width 
but  rather  a  greater  discrimination  in  the  sorting  of  the  catch. 


The  Concepts  of  Structure  and  Function        29 

not  more  so  than  is  the  use  of  pure  materials  in  a  chemical 
laboratory. 

The  state  of  affairs  in  anthropology  is  perhaps  not  quite 
as  bad  as  I  have  drawn  it.  We  have  on  the  one  hand  Radcliffe- 
Brown,  who  tends  to  see  all  the  elements  of  culture  in  their 
bearing  upon  the  solidarity,  the  integration,  of  the  group. 
He  insists  that  purely  psychological  considerations  are  irrele- 
vant in  this  matter  and  he  discriminates  social  structure  from 
pragmatic  function  and  regards  each  as  presenting  its  special 
type  of  problem  to  be  studied  by  its  own  special  discipline. 
And  on  the  other  hand  we  have  the  developing  science  of 
economics  which  specialises  in  that  type  of  social  function 
which  is  concerned  with  the  satisfaction  of  human  needs  and 
desires — a  worthy  methodological  attempt  perhaps  hampered 
by  a  narrow  view  of  human  needs. ^ 

These  are  considerable  advances  and  in  this  book  I  shall 
build  upon  the  disciplines  discriminated  by  Radcliffe- Brown. 
I  shall  endeavour  to  consider  the  functional  position  (using 
the  term  in  its  widest  philosophical  sense)  of  the  naven  cere- 
monies from  five  different  points  of  view.  That  is  to  say, 
I  shall  classify  functions  into  five  categories  so  that  the  various 
parts  of  the  book  will  illustrate  five  different  methods  of 
approach  to  the  problems  of  culture  and  society.  As  far  as 
possible,  I  shall  keep  these  methods  of  approach  separate  and 
shall  indicate  where  I  depart  from  this  procedure. 

The  categories  which  I  have  chosen  are: 

1 .  Structural  or  "  logical "  relationships,  between  the  cogni- 
tive aspects  of  the  various  details  of  cultural  behaviour: 
the  cognitive  reasons  for  behaviour. 

2.  Affective^   relationships,    between   details   of  cultural 

^  Sargent  Florence  {Economics  and  Human  Behaviour,  chap,  iv,  1927) 
discusses  the  position  taken  up  by  Henderson  in  economics — the  assump- 
tion that  we  can  retain  "  calculating  man  "  as  a  basic  postulate  of  economics 
provided  we  make  no  assumptions  as  to  the  terms  in  which  the  calcula- 
tion takes  place.  This  restriction  provides  a  convenient  methodological 
isolation  of  economics  from  psychology. 

^  Although  the  terms  "affect"  and  "cognition"  are  now  in  some  dis- 
favour among  psychologists,  I  have  been  compelled  to  adopt  the  adjectives 
derived  from  them.  I  believe  that  the  dangers  and  fallacies  inherent  in 
these  terms  may  to  a  great  extent  be  avoided  if  we  speak  not  of  "  affective 
mechanisms"  and  "  cognitive  mechanisms"  but  of  affective  and  cognitive 
aspects  of  the  behaviour  of  a  single  mechanism,  the  individual. 


30         The  Concepts  of  Structure  and  Function 

behaviour  and  the  basic  or  derived  emotional  needs  and 
desires  of  the  individuals :  the  affective  motivation  of  details 
of  behaviour. 

3.  Ethological  relationships,  between  the  emotional  aspects 
of  details  of  cultural  behaviour  and  the  emotional  emphases 
of  the  culture  as  a  whole. 

4.  Eidological  relationships,  between  the  cognitive  aspects 
of  details  of  cultural  behaviour  and  the  general  patterning 
of  the  cultural  structure. 

5 .  Sociological  relationships,  between  the  cultural  behaviour 
of  the  individuals  and  the  needs  of  the  group  as  a  whole :  the 
maintenance  of  solidarity,  etc. 

Of  these  five  categories  of  relationship  between  bits  of 
anthropological  material,  it  is  clear  that  one,  the  sociological, 
is  in  a  class  apart.  Of  the  others,  the  ethological  is  very  closely 
related  to  the  affective,  and  the  eidological  is  very  closely 
related  in  an  analogous  manner  to  the  structural.^ 

Besides  these,  there  are  two  other  methods  of  approach 
which  I  shall  not  consider  in  any  detail,  namely,  those 
of  developmental  psychology  and  economics.  These  are 
omitted,  not  because  I  decry  their  importance  but  because 
I  did  not  collect  in  the  field  the  material  which  would  have 
illustrated  these  approaches. 

Of  these  categories  but  little  need  be  said  at  this  stage  in 
our  enquiry.  The  concepts  of  structure,  both  social  and  cul- 
tural, I  owe  very  largely  to  Radcliffe- Brown;  while  another 
concept,  that  of  "pragmatic  function",  was  derived  from 
Malinowski.  This  latter  concept  I  have  subdivided  for 
several  reasons. 

Let  us  consider  some  rather  more  concrete  examples  of 
"pragmatic"  functioning.  The  term  is  applied  to  such  effects 
as :  increase  of  sociability  among  individuals ;  increased  soli- 
darity of  the  community;  increased  family  pride;  the  con- 
firmation of  the  privileges  of  individuals  and  the  enforcing 
of  their  rights  and  duties ;  substantiation  of  belief  in  magical 

^  Indeed  so  close  are  the  resemblances  between  these  pairs  of  categories 
that  I  have  been  led  to  refer  to  the  affective  functions  of  behaviour  as  "the 
expression  of  ethos  in  behaviour";  and  I  might  similarly  have  referred 
to  the  structural  functions  as  "the  expression  of  eidos  in  behaviour". 


The  Concepts  of  Structure  and  Function        3 1 

efficiency  and  the  strengthening  of  traditional  law  and  order.  ^ 
At  a  first  glance  many  of  these  functions  would  appear  to  be 
"useful",  and  anthropologists  have  been  tempted  to  attach 
value  to  functions  of  this  sort  without  more  ado.  But  their 
usefulness  becomes  more  problematical  when  we  realise  that 
many  of  these  effects  may  be  mutually  antagonistic,  that  for 
example  an  increase  in  family  pride  may  well  disrupt  the 
solidarity  of  the  community;  or,  again,  excessive  emphasis 
upon  solidarity  within  the  community  may  conceivably  lead 
to  wars  with  outside  peoples  and  so  to  the  destruction  of  the 
status  quo.  In  the  light  of  the  potential  antagonism  between 
these  effects,  it  would  seem  that  value  can  only  be  attributed 
to  them  with  great  caution ;  and  in  a  sample  list  of  valuable 
effects  we  should  certainly  include  all  the  opposites  of  the 
effects  stressed  by  the  Functional  School — the  reduction  of 
solidarity,  the  reduction  of  family  pride,  the  relaxing  of 
duties,  etc. 

Again,  if  we  consider  this  sample  list  of  functions  in  diverse 
settings,  their  value  becomes  still  more  problematic.  A  nomad 
people  living  on  the  barest  margin  of  physical  sustenance, 
and  drastically  regulating  population  by  infanticide,  is  not 
likely  to  be  interested  in  integrating  a  larger  population.  The 
truth  of  the  matter  is  that  different  peoples  attach  very 
different  values  to  these  various  effects.  Some  peoples  ap- 
prove of  sociability  while  others  are  inclined  to  frown  upon 
the  sociable  talkative  individual  and  to  regard  him  as  un- 
dignified and  intruding.  The  "strength  of  traditional  law" 
is  very  important  in  some  African  communities,  but  the 
phrase  is  almost  meaningless  when  applied  to  the  latmul  who 
have  a  highly  individualistic  culture  and  will  readily  respect 
the  law-breaker  if  he  have  but  sufficient  force  of  personality. 
The  latmul  too — like  the  Irish — set  no  exaggerated  value 
upon  extremely  orderly  behaviour.  They  desire  a  degree  of 
orderliness  very  different  from  that  demanded  by,  say,  a 
Quaker  community.  And  so  on;  whichever  of  these  effects 
we  take  we  shall  find  that  among  certain  peoples  that  effect 

^  This  sample  list  of  functions,  and  the  term  "pragmatic  function" 
applied  to  them,  are  derived  from  Malinowski's  article  "Anthropology", 
Supplement  to  the  Encyclopaedia  Britannica,  1926.  They  are  the  functions 
which  he  ascribes  to  myths  of  various  types. 


32         The  Concepts  of  Structure  and  Function 

is  highly  valued  while  among  others  it  is  ignored  or  despised. 
The  same  conclusions  apply  not  only  to  such  abstract  effects 
as  the  strength  of  law  and  integration,  but  even  to  such  con- 
crete matters  as  the  food  supply,  one  people  setting  a  high 
value  on  good  living  while  an  ascetic  group  may  take  but 
little  interest  in  the  flesh  pots. 

It  is,  I  think,  time  that  anthropologists  took  account  of 
this  enormous  variation  in  the  value  which  different  peoples 
set  upon  the  effects  which  modern  functional  anthropology 
has  to  offer  them.  In  the  light  of  this  variation  it  is  obviously 
dangerous  to  state  that  the  significant  pragmatic  function  of 
a  given  detail  of  a  given  culture  is  the  increase  of  family 
pride — unless  it  can  first  be  shown  that  family  pride  is  really 
one  of  the  effects  which  is  valued  in  other  contexts  of  the 
particular  culture  which  we  are  studying. 

Here,  then,  we  arrive  at  a  very  different  type  of  functional 
relationship.  We  began  by  considering  the  relation  between 
details  of  culture  on  the  one  hand  and  the  needs  and  desires 
of  the  individual  on  the  other.  But  we  are  now  faced  with 
the  fact  that  we  cannot  guess  at  those  needs  and  desires  but 
must  first  deduce  them  from  the  emphases  of  the  culture 
as  a  whole.  Thus,  if  we  isolate  from  "pragmatic  function" 
the  concept  of  affective  function  which  we  may  define,  rigidly, 
in  terms  of  the  relation  between  details  of  culture  and  the 
emotional  needs  of  individuals,  it  follows  that  we  must  con- 
struct another  category  for  the  relationship  between  the 
emotional  content  of  the  particular  detail  of  behaviour  whose 
functions  we  are  studying  and  the  emotional  emphases  of 
the  culture  as  a  whole.  This  category  of  function  I  shall  call 
ethological ;  and  I  shall  use  the  word  ethos  to  refer  collectively 
to  the  emotional  emphases  of  the  culture. 

Similar  considerations  (see  footnote  to  the  word  "  logical ", 
p.  25)  will  lead  us  from  the  isolation  of  the  concept  of  cul- 
tural structure  to  the  isolation  of  eidos.  From  the  examination 
of  the  premises  in  the  structure  of  a  particular  culture  we  can 
fit  them  together  into  a  coherent  system  and  finally  arrive 
at  some  general  picture  of  the  cognitive  processes  involved. 
This  general  picture  I  shall  call  the  eidos  of  the  culture. 

But,  though  the  relationship  between  eidos  and  structural 


The  Concepts  of  Structure  and  Function        33 

premises  is  so  closely  analogous  to  that  between  ethos  and 
affective  functions,  there  is  one  anomaly  to  which  attention 
must  be  called.  We  have  seen  that  the  study  of  ethos  is  a 
necessary  preliminary  to  any  conclusions  about  pragmatic 
functions.  But  in  the  case  of  structure  and  eidos,  the  order 
of  procedure  is  reversed.  The  details  of  the  cultural  structure 
are  to  be  first  studied  and  from  these  the  eidos  is  deduced. 
This  reversal  is  due  to  the  fact  that  in  studying  cultural 
structure  we  are  concerned  with  the  manifest  cognitive  con- 
tent of  behaviour,  whereas  in  studying  pragmatic  functions 
we  are  concerned  with  the  much  more  obscure  emotional 
content.  The  manifest  content  can  be  described  piece  by  piece 
and  the  underlying  system  deduced  from  the  resulting  de- 
scription. But  emotional  significance  can  only  be  ascribed 
after  the  culture  as  a  whole  has  been  examined. 

Apart  from  this  diflFerence  in  procedure,  the  ethological 
and  eidological  approaches  to  culture  are  very  closely  analo- 
gous. Both  are  based  upon  the  same  fundamental  double 
hypothesis:  that  the  individuals  in  a  community  are  stan- 
dardised by  their  culture ;  while  the  pervading  general  char- 
acteristics of  a  culture,  those  characteristics  which  may  be 
recognised  over  and  over  again  in  its  most  diverse  contexts, 
are  an  expression  of  this  standardisation.  This  hypothesis  is, 
in  a  sense,  circular ;  it  is  supposed  that  the  pervading  charac- 
teristics of  the  culture  not  only  express,  but  also  promote  the 
standardisation  of  the  individuals. 

For  this  double  assumption,  we  are  chiefly  indebted  to 
Dr  Benedict ;  and  the  concepts,  ethos  and  eidos,  which  I  am 
suggesting  may  be  regarded  as  subdivisions  of  her  more 
general  concept.  Configuration.  The  eidos  of  a  culture  is  an 
expression  of  the  standardised  cognitive  aspects  of  the  indi- 
viduals, while  the  ethos  is  the  corresponding  expression  of 
their  standardised  affective  aspects.  The  sum  of  ethos  and 
eidos,  plus  such  general  characteristics  of  a  culture  as  may  be 
due  to  other  types  of  standardisation,  together  make  up  the 
configuration.^ 

^  Dr  Benedict  was  kind  enough  to  read  a  large  part  of  this  book  in 
manuscript  form  before  it  was  completed  and  she  agrees  with  me  that 
this  is  a  correct  statement  of  the  relationship  of  ethos  and  eidos  to  con- 
figuration. 


34         The  Concepts  of  Structure  and  Function 

This  hypothesis  of  the  standardisation  of  the  individuals 
goes  a  long  way  to  make  clear  the  inter-relationships  of  the 
various  disciplines  which  we  have  so  far  isolated — the  struc- 
tural and  eidological,  pragmatic  and  ethological — but  it  also 
brings  to  the  fore  another  separable  discipline,  another 
category  of  function.  This  discipline  is  generally  called 
Developmental  Psychology,  the  study  of  the  moulding  of  the 
individual  by  the  circumstances  of  his  environment,  the 
psychological  effects  upon  him  of  the  impact  of  culture. 

Thus  we  have  worked  down  from  the  study  of  cultural 
processes  to  the  study  of  individual  psychology,  linking  the 
two  ends  of  the  scale  by  the  abstract^  concept  of  the  "stan- 
dardised individual ". 

Returning  now  to  the  raw  material  of  anthropology,  the 
facts  of  native  behaviour  noted  down  by  the  field  worker, 
we  may  see  the  whole  of  this  mass  as  relevant  to  a  final 
separable  discipline.  This  is  concerned  with  the  effects  of 
behaviour  in  satisfying  the  needs  of  groups  of  individuals,  in 
maintaining  or  impeding  the  continued  existence  of  that 
abstract  organism — the  Society.  This  category  of  function 
I  shall  call  Social  Function  and  shall  use  the  term  Sociology 
in  a  restricted  sense  to  refer  to  its  study,  as  distinct  from  that 
of  the  other  separable  disciplines, ^  In  this  discipline  of 
sociology,  as  Radcliffe-Brown  has  pointed  out,  individual 
psychology  becomes  irrelevant  to  the  same  extent  that  Atomic 
Physics  is  methodologically  irrelevant  in  the  study  of  Bio- 
chemistry. 

^  It  may  be  possible  in  the  light  of  future  work  to  define  the  "stan- 
dardised individual"  in  rather  more  concrete  terms.  For  the  present  it  is 
not  possible  to  say  whether  by  this  concept  we  mean  the  "average"  or 
the  "  ideal "  individual.  Nor  is  it  at  all  clear  what  differences  will  be  intro- 
duced into  the  theory  of  culture  according  to  our  interpretation  of  the 
"standard".  These  matters  can  only  be  cleared  up  after  more  intensive 
work  has  been  done  on  normal  and  maladjusted  individuals  in  various 
cultures. 

^  Malinowski,  I  believe,  includes  under  the  term  "pragmatic  function" 
not  only  the  type  of  function  to  which  I  apply  that  term  but  also  what  I  am 
here  calling  "social  function".  The  difference  in  our  points  of  view 
springs  largely  from  his  definition  of  function  in  terms  of  the  direct 
or  indirect  satisfaction  of  the  individuals.  I  have  defined  pragmatic  func- 
tion in  terms  of  direct  satisfaction  only,  leaving  the  various  steps  in  the 
roundabout  routes  of  indirect  satisfaction  for  separate  examination  in 
the  other  categories  of  function. 


CHAPTER  IV 

Cultural  Premises  relevant  to  the 
Wau-laua  Relationship 

IN  this  section  and  that  which  follows  it,  I  shall  view  the 
elements  of  latmul  culture  as  links  in  a  structural  system. 
I  shall  see  the  mother's  gifts  of  food  to  her  child  and  the 
child's  loyalty  to  its  mother  as  details  of  behaviour  which  are 
assumed  and  expected  in  this  culture,  as  premises.  Later,  in 
another  section,  I  shall  consider  such  cultural  facts  in  their 
relation  to  needs  and  desires.  I  shall  for  the  present  endeavour 
only  to  show  that  the  premises  of  latmul  culture  are  connected 
together  into  a  coherent,  "logical"  system.  I  shall  consider 
syntheses  of  the  type:  a  mother  gives  food  to  her  child. 
A  mother's  brother  is  identified  with  the  mother.  Therefore 
a  mother's  brother  gives  food  to  his  sister's  son. 

For  the  handling  of  cultural  elements  as  parts  of  a  structure 
Radcliffe- Brown  has  suggested  a  very  useful  technique  in- 
volving the  use  of  the  technical  term  identification.  This  term 
has  given  rise  to  a  certain  amount  of  misunderstanding  and 
I  must  explain  the  sense  in  which  I  use  it.  The  term  is  a 
shorthand  method  of  referring  collectively  to  the  structural 
aspects  of  large  numbers  of  details  of  a  culture.  When  I  say 
that  siblings  are  identified  I  mean  that  in  this  culture  there  is 
a  large  number  of  details  of  culturally  standardised  behaviour 
in  which  the  behaviour  of  one  sibling  resembles  that  of  the 
other;  and  I  also  mean  that  many  details  of  the  behaviour 
of  outside  individuals  towards  one  sibling  are  reproduced  in 
their  behaviour  towards  the  other.  Further,  we  shall  see 
that,  in  latmul  culture  at  least,  when  two  individuals  are 
identified  it  is  usual  for  one  of  them  to  behave  as  if  he  had 
performed  certain  acts  which  as  a  matter  of  fact  were  per- 
formed by  the  other. ^    Similarly,  outside  persons  behave 

^  The  most  striking  case  of  this  occurs  in  the  identification  between 
Hving  individuals  and  their  ancestors.  Among  the  latmul  one  often  hears 
such  statements  as  "/  was  on  the  stern  of  the  canoe  which  brought  the 
first  people  to  Mindimbit  village!",  meaning  that  an  ancestor  who  had 
the  same  name  as  the  speaker  played  this  part  in  the  founding  of  the 
village. 


36  Cultural  Premises  relevant  to 

towards  the  one  individual  as  if  he  had  performed  the  acts 
which  were  performed  by  the  other. 

The  statement  that  siblings  are  identified  does  not  imply 
the  absence  of  details  of  behaviour  which  would  discriminate 
between  them.  Thus  the  word  identification  is  used  relatively 
and  not  absolutely.  It  is  common  to  find  that  in  one  series 
of  contexts  two  individuals  are  identified,  while  in  another 
series  they  are  contrasted.  Thus  we  shall  see  that,  in  latmul 
culture,  father  and  son  are  identified  in  their  relations  with 
the  outside  world  while  between  themselves  they  are  dras- 
tically contrasted.  In  so  far  as  they  are  identified,  they  are 
regarded  as  allies  but  in  so  far  as  they  are  contrasted,  they 
are  regarded  as  mutually  opposed  in  their  interests. 

The  technique  of  describing  identification  is  by  far  the 
best  so  far  evolved  for  the  mapping  of  kinship  systems ;  and 
for  the  study  of  naven  we  have  to  consider  one  important 
discrimination  and  four  major  types  of  identification : 

I.  That  mother's  brother  {wau)  is  distinct  from  father's 
brother  or  father  {iiyai').  This  discrimination  is  made  overtly 
in  the  kinship  terms  and  is  implicit  in  many  of  the  patterns 
of  behaviour  towards  these  relatives.  In  the  present  context 
this  premise  may  be  suitably  illustrated  by  a  myth  about 
naven : 

A  boy  and  girl — brother  and  sister — went  out  to  work  sago  in  a 
sago  swamp.  As  they  went  in  their  canoe  the  mayflies  came  out 
and  the  surface  of  the  river  was  covered  with  them,  so  they  set  to 
work  and  collected  a  great  quantity  of  mayflies.  The  fish  too  began 
to  rise  and  the,.boy  speared  a  number  of  fish.  Then  they  went  on 
to  the  sago  swamp  where  the  boy  felled  a  sago  palm  and  beat  the 
pith  out  of  the  trunk.  The  girl  took  the  pith  and  washed  it  and 
collected  the  sago.  On  their  way  home  they  were  met  by  their 
father  and  mother  who  had  come  out  to  meet  them  in  a  canoe, 
and  who  exclaimed:  "Who  speared  the  fish?"  and  "Who  beat  the 
sago?"  The  boy  and  girl  replied  that  they  two  had  speared  the 
fish  and  beaten  the  sago.  Then  the  father  took  off  his  pubic  apron 
and  the  mother  took  off  her  skirt  and  they  began  to  dance  a  naven. 
The  boy  and  girl  exclaimed:  "Why  are  you  dancing?"  and  were 
utterly  ashamed.  So  the  boy  and  girl  put  the  sago  and  the  fish 
back  in  their  canoe  and  paddled  away  to  their  wau.  When  they 
reached  the  house,  wau  and  mbora  (mother's  brother's  wife)  came 
out  and  took  up  the  fish  and  sago,  and  wau  took  off  his  apron  and 
mbora  took  off  her  skirt  and  danced  a  naven,  and  mbora  took  up 


the  Wau-Iaua  Relationship  37 

the  fish  and  sago  and  cooked  them  and  she  and  wau  ate.  Later 
mbora  brought  a  pig  which  she  had  fattened  and  wau  presented  it 
to  the  two  laua  who  set  their  feet  upon  it.  Then  they  went  and 
prepared  shell  valuables  and  presented  them  to  the  wau. 

This  discrimination  between  father  and  mother's  brother 
is  especially  conspicuous  in  all  those  contexts  in  which  their 
behaviour  is  influenced  by  their  membership  in  different 
patrilineal  clans.  Every  clan  has  its  series  of  myths  about 
ancestors.  These  myths  and  the  spells  which  are  based  upon 
them  are  secrets  which  are  carefully  preserved  by  the  descen- 
dant clans.  A  man  will  normally  pass  on  his  esoteric  know- 
ledge to  one  of  his  sons ;  but  sometimes  he  will  impart  the 
clan  secrets  and  spells  to  a  sister's  son.  In  the  latter  case  the 
esoterica  must  be  ceremonially  paid  for  with  shell  valuables 
and  the  sister's  son  is  supposed  not  to  pass  on  the  wau's 
esoterica  to  his  own  sons. 

The  entrusting  of  esoterica  to  the  laua  is  a  strong  affirma- 
tion of  the  tie  between  him  and  his  zvau.  It  implies  con- 
siderable trust  and  confidence  on  the  waus  part  since  the 
laua,  when  he  has  been  told  the  secrets,  may  easily  use  them 
in  debate  to  his  wau's  undoing  and  shame.  The  motive  which 
leads  the  wau  to  take  this  risk  is,  I  think,  a  hope  that  the  laua 
will  speak  on  his  side  in  debates.  Malikindjin,  who  spurned 
the  idea  that  he  needed  any  assistance  in  debate  (cf.  p.  165), 
yet  taught  a  part  of  his  esoteric  learning  to  his  laua,  saying, 
"When  I  am  dead,  my  laua  will  help  my  son  in  his  debates". 
I  was  told  of  one  case  only  in  which  a  laua  used,  in  debate 
against  his  wau,  information  which  the  latter  had  confided 
to  him. 

Sometimes  a  clan  has  the  misfortune  to  lose  its  ancestral 
traditions  and  must  then  buy  them  back  from  its  lauas. 

An  interesting  case  of  discrimination  between  wau  and 
father  occurs  in  Mindimbit,^  in  the  procedure  which  accom- 
panies a  small  boy's  first  killing  of  a  human  being.  For  this 
important  occasion  the  father  of  the  boy  will  obtain,  either 
by  capture  or  purchase,  a  captive  from  a  neighbouring  com- 

^  This  ceremonial  point  is  apparently  not  made  in  Kankanamun  and 
Palimbai,  where  I  asked  my  informants  about  the  corresponding  pro- 
cedure. They  seemed  to  think  it  did  not  matter  who  helped  the  boy  in 
his  kill :  that  probably  the  father  would  do  it. 


38  Cultural  Premises  relevant  to 

munity  and  this  captive  will  be  speared  by  the  boy.  But  the 
killing  is  in  many  cases  carried  out  while  the  child  is  still  too 
young  to  wield  a  spear ;  and  so  he  must  be  helped  in  the  per- 
formance of  the  deed.  This  help  must  not  be  provided  by 
the  father  because  the  natives  say  that  the  kill  would  in  that 
case  be  the  father's  achievement  and  not  that  of  the  son. 
But  the  wau  can  help  the  child;  he  will  hold  the  spear  and 
direct  the  child's  thrust,  and  after  this  the  kill  will  count  as 
the  achievement  of  the  child. 

Similarly  wau  is  discriminated  from  father  in  economic 
contexts.  Here  the  pattern  is  the  inverse  of  that  in  contexts 
of  achievement.  When  a  boy  is  collecting  valuables  to  provide 
a  bride  price,  his  father  will,  as  a  matter  of  course,  help  him 
and  will  not  expect  any  formal  recompense  for  the  valuables 
which  he  contributes.  The  wau  hiay  also  contribute,  but  in 
this  case  he  will  expect  some  return.  Usually  the  boy  will 
make  this  return  in  the  form  of  a  contribution  to  the  wau's 
son  when  the  latter  is  collecting  valuables  to  get  a  wife.  Thus, 
in  contexts  which  concern  valuables,  the  boy  is  grouped  with 
his  father  and  separated  from  the  maternal  clan;  while  in 
contexts  of  achievement  the  boy  is  united  with  the  maternal 
clan  and  separated  from  his  father.  But  in  both  types  of 
context  the  wau  is  discriminated  from  the  father. 

In  spite  of  this  discrimination,  we  shall  see  that  there  is 
also  a  trace  of  identification  between  father  and  wau  and  that 
this  appears  even  in  some  details  of  their  behaviour  towards 
the  laua  (cf.  p.  83).  In  other  contexts,  in  which  the  same 
two  persons  are  concerned,  not  dealing  with  the  laua  to 
whom  they  appear  as  wau  and  nyai,  but  dealing  with  the 
outside  world  to  whom  they  appear  as  a  pair  of  brothers- 
in-law,  we  shall  see  that  co-operation  and  competition,  identi- 
fication and  discrimination,  are  very  evenly  mixed  in  the 
patterning  of  their  mutual  relationship. 

2.  That  there  is  a  certain  identity  between  a  father  and 
his  son.  This  premise  is  expres?ed,  for  example,  in  the 
patrilineal  system  of  clans,  and  in  many  other  ways.  For  the 
present  purpose  it  will  suffice  if  we  examine  in  some  detail 
only  one  of  the  contributory  lines  of  evidence — that  from 
the  vocabulary  of  kinship — and  omit  the  others.    I  have 


the  Wau-laua  Relationship  39 

already  published  ^  a  series  of  examples  of  the  latmul  ten- 
dency to  summarise  and  formulate  the  behaviour  of  indi- 
viduals by  using  the  terminology  of  kinship  in  an  analogic 
manner.  This  usage  is  sufficiently  familiar  in  our  own  society 
to  need  no  stressing,  and  since  the  latmul  have  carried  this 
practice  further  than  we  have,  we  can  with  unusual  safety 
refer  to  the  vocabulary  as  evidence  of  the  formulated  struc- 
ture of  the  kinship  system. 

The  vocabulary  of  kinship  contains  a  number  of  phrases 
in  which  two  kinship  terms  are  used  in  juxtaposition  to  indi- 
cate a  group  of  relatives  who  are  closely  associated  in  the 
native  mind,  and  from  such  compound  forms  we  may  deduce 
a  close  identification  between  the  behaviour  of  and  towards 
the  types  of  kin  who  are  actually  named  in  the  compound 
phrases.  As  an  example  we  may  take  the  term  nggwail- 
warangka  in  which  the  word  figgwail  means  father's  father 
and  warangka  means  father's  father's  father.  The  whole 
compound  is  a  collective  term  for  patrilineal  ancestors,  a  group 
of  relatives  who  are  classified  together  in  latmul  culture.^ 

There  are  several  other  cases  in  which  compound  terms 
meaning  "a  certain  relative  and  his  father"  are  used  to 
denote  whole  groups  of  relatives  who  are  identified  together, 
for  example:  wau-mbuambo  (mother's  brother  and  mother's 
brother's  son),  which  is  a  collective  term  for  all  the  members 
of  the  maternal  clan ;  towa-naisagut  (abbreviated  from  tawontu- 
naisaguty  and  meaning  literally  wife's  brother  and  wife's 
father),  which  is  a  collective  term  for  the  whole  of  the  wife's 
clan;  laua-ianan  (sister's  son  and  sister's  son's  son),  a  col- 
lective term  for  the  descendants  of  the  men  who  married 
the  speaker's  sister.  In  all  these  terms  the  identity  of  a  son 
with  his  father  is  emphasised. 

A  similar  identification  is  implied  by  an  incorrect  use  of 
kinship  terms  for  the  mother's  brother's  son  and  mother's 
brother's  son's  wife.  These  two  relatives  are  normally  spoken 
of  as  mbuambo,  a  term  which  identifies  them  with  mother's 
father  and  mother's  mother.   But  one  informant  in  Palimbai 

^  Oceania,  1932,  p.  266. 

^  The  same  set  of  relatives  is  constantly  invoked  in  name-songs,  in 
the  refrain:  " nyai'  nya!  nyamun  a!" — (O  Father!  O  Elder  brother!). 


4©  Cultural  Premises  relevant  to 

told  me  that  after  the  death  of  the  mother's  brother,  the  term 
wau  is  used  for  the  mother's  brother's  son,  and  the  term 
mbora  (mother's  brother's  wife)  is  then  appHed  to  the  mother's 
brother's  son's  wife.  Other  informants  said  that  this  was 
incorrect,  but  I  suspect  that  if  the  culture  were  left  un- 
disturbed this  usage  might  become  more  general.  Among 
the  neighbouring  Tshuosh  people  there  is  no  distinction  in 
terminology  between  the  male  members  of  different  genera- 
tions of  the  maternal  clan;  all  are  called  wau.  Among  the 
latmul  too,  such  a  use  of  terms  would  fit  with  the  wau's 
motive  in  teaching  his  esoterica  to  his  laua — "When  I 
am  dead  my  laua  will  help  my  sons  in  their  debates" 
(cf.  p.  37). 

Here  it  must  be  stressed  that  the  identification  of  father 
and  son,  while  it  is  conspicuous  in  their  relations  with  out- 
side persons,  is  not  allowed  to  become  a  basis  for  easy 
intimacy  between  them.  Such  intimacy  is  sternly  discouraged, 
and  in  its  place  the  conflict  between  successive  generations 
is  emphasised  and  is  handled  in  a  dignified  manner.  While 
the  father  may  resent  bitterly  and  avenge  any  scolding  or 
blows  which  others  may  administer  to  his  son,  he  himself  is 
in  a  disciplinary  position  and  may  strike  his  son  in  anger. 
But  in  contrast  to  the  disciplinary  position  of  the  father,  we 
find  that  the  son  will  gradually  displace  the  father.  These  two 
facts  are  the  basis  of  the  formulations  which  govern  their 
relationship.  The  possibility  of  any  overlapping  between 
father  and  son,  in  social  status,  is  drastically  limited  by  taboos 
which  are  felt  not  to  be  arbitrary  but  rather  to  indicate  the 
only  proper  way  of  meeting  the  situation.  The  father  must 
never  profit  at  the  expense  of  his  son.  He  will  never  eat 
anything  which  the  son  has  grown  in  his  garden,  nor  may 
he  eat  food  gathered  by  his  son,  even  though  it  be  from  the 
father's  own  garden.  When  they  are  in  a  canoe  together,  the 
son  normally  takes  the  position  of  honour  in  the  bow  and, 
while  he  is  there,  the  father  will  never  scoop  up  water  from 
alongside  to  quench  his  thirst.  In  the  initiatory  system  the 
son  is  normally  two  grades  below  his  father  (see  diagram, 
p.  245).  But  the  people  are  continually  readjusting  the  balance 
between  the  grades  and  it  occasionally  happens  that  an  attempt 


the  Wau-laua  Relationship  41 

is  made  to  promote  a  boy  into  the  grade  of  which  his  father 
is  a  member.  Such  an  attempt  is  violently  resisted  by  the 
boy,  who  feels  that  this  will  be  to  intrude  upon  his  father's 
dignity.  But  the  father  will  make  no  resistance  to  his  son's 
promotion,  since  it  ill  becomes  him  to  defend  his  own  pride 
at  the  price  of  his  son's  advancement.  I  think  that  in  most 
cases  the  promotion  is  finally  effected.  But  in  one  case  which 
I  witnessed  it  was  successfully  resisted  by  the  son  after  he 
had  finally  said  that  he  would  abide  by  his  wau's  decision. 
The  wau  was  consulted  and  opposed  the  promotion.  Thus 
the  son  is  as  much  concerned  as  the  father  to  avoid  any  over- 
lapping of  social  status. 

The  taboos  which  a  father  observes  are  obeyed  chiefly 
because  the  father  feels  them  to,  be  right.  But  they  are  to 
some  extent  reinforced  by  a  belief  that  if  he  eats  food  col- 
lected by  his  son  he  will  soon  grow  old — perhaps  a  symbolic 
statement  of  the  fact  that  in  so  doing  the  father  will  be 
behaving  like  a  grandfather.^  But  more  important  than  this 
sanction  is  the  strength  of  public  feeling.  The  idea  of  in- 
timacy between  father  and  grown  son  is  shocking  to  the 
latmul.  While  I  was  on  the  Sepik,  a  European  father  and 
son  settled  together  on  the  Ceram  River  many  miles  away, 
where  they  started  a  small,  isolated  tobacco  plantation.  The 
latmul  heard  of  this  and  were  surprised  and  shocked.  They 
came  to  me  to  ask  if  the  two  were  really  father  and  son  and 
were  disgusted  when  I  told  them  that  it  was  so.  They  said, 
"  Has  the  father  no  shame? " 

These  taboos  have  very  little  classificatory  spread.  They 
apply  only  to  own  father,  father's  own  younger  brothers 
(father's  elder  brother  is  addressed  as  mbuambo,  a  term 
which  means,  among  other  things,  "maternal  grandfather") 
and  own  father's  tambinyen  or  partner  in  the  opposite 
moiety.2 

The  behaviour  patterns  between  father  and  son  are  marked 

^  In  latmul  social  organisation,  alternate  generations  in  the  patri- 
lineal line  are  grouped  into  two  opposite  tnbapma  (literally,  "lines")  as 
in  Australian  kinship  systems.  It  follows  from  this  that  a  father,  in 
identifying  himself  with  his  son,  would  be  behaving  as  if  he  were  the 
boy's  grandfather. 

^  See  Oceania,  1932,  p.  264. 


42  Cultural  Premises  relevant  to 

typically  by  restraint  on  both  sides.  Each  respects  the  other, 
and  both  are  rather  quiet  and  stiff  when  talking  together. 
But  this  does  not  mean  that  an  outsider  who  is  speaking  in 
the  presence  of  one  of  them  can  make  remarks  about  the 
other.  A  very  real  identification  of  interests  and  of  needs 
exists  between  them  in  spite  of  their  mutual  avoidance  of 
overlapping  status. 

3.  That  in  spite  of  the  patrilineal  system,  the  child  is  linked 
in  some  way  with  its  mother's  clan.  This  premise  may  be 
recognised,  symbolically  expressed,  in  the  native  theory  of 
gestation.  It  is  supposed  that  the  bones  of  the  child  are  a 
product  of  the  father's  semen,  while  its  flesh  and  blood 
(somewhat  less  important)  are  provided  by  the  mother's 
menstrual  blood.  This  idea  is  carried  logically  to  the  con- 
clusion that  the  afterbirth,  lacking  bones,  is  therefore  the 
child  of  the  mother  only. 

The  child  is  in  one  way  a  member  of  its  father's  clan,  but 
at  the  same  time  it  is  in  quite  a  different  way  a  member  of 
its  mother's  clan.  This  formulation  is  expressed  by  the  giving 
of  two  sets  of  names  to  the  child.  One  set  are  names  of  the 
totemic  ancestors  of  the  father's  clan,  while  the  other  set  are 
names  of  the  totemic  ancestors  of  the  mother's  clan.  These 
two  sets  of  names  have  different  terminations :  the  patrilineal 
names  end  in  suffixes  which  mean  "man",  "woman", 
"  body  ",  etc.,  while  the  names  of  the  ancestors  of  the  mother's 
clan  end  in  the  suffix,  -awan,  which  probably  means  "mask" 
and  is  perhaps  connected  with  the  custom  mentioned  above 
of  dancing  in  masks  which  represent  the  totemic  ancestors 
of  the  maternal  clan.^ 

These  two  sets  of  names  are  apparently  felt  to  represent 
two  quite  different  facets  of  the  individual's  personality. 
After  death  a  man  dwells  in  the  land  of  the  dead  under  the 
names  which  his  wau  gave  him,  but  he  is,  at  the  same  time, 
reincarnated  in  his  son's  son  under  his  patrilineal  name.  In 
the  techniques  of  black  magic  it  is  the  -mvan  name  which  is 

^  The  etymology  of  suffixes  attached  to  the  names  which  are  given 
by  the  ivau  is  doubtful.  The  corresponding  names  given  to  female  children 
end  in  the  suffix,  -yelishi.  This  word  means  "old  woman"  or  "old  lady". 
It  is  possible  that  the  word  awan  means  "old  man"  and  that  this  word  is 
only  applied  metaphorically  to  the  masks  called  awan. 


the  Wau-laua  Relationship  43 

used  by  the  sorcerer  in  addressing  the  figure  which  represents 
his  victim;  and  the  people  take  some  sHght  care  not  to  use 
the  -awan  names  after  dark,  lest  the  witches  overhear  and 
use  the  name  to  the  detriment  of  its  owner.^ 

In  general,  it  would  seem  that  the  names  given  by  the  wau 
represent  a  more  mysterious  aspect  of  the  personality  than 
those  given  by  the  father.  In  a  sense  the  link  with  the 
maternal  clan  is  analogous  to  that  with  the  paternal  and  each 
is  expressed  by  the  giving  of  names ;  but  while  according  to 
the  patrilineal  name  the  child  is  a  "man"  or  "body",  ac- 
cording to  the  matrilineal  he  is  a  "mask".  Similarly  in  the 
patrilineal  theory  of  reincarnation  the  child  is  concretely 
stated  to  be  a  reincarnation  of  a  father's  father,  while  his 
identity  with  the  ancestors  of  his  mother's  clan  is  only 
symbolically  expressed  in  ritual  and  ceremonial  behaviour. 
This  ritual  behaviour  may  conveniently  be  enumerated  at 
this  point: 

{a)  The  giving  of  -awan  and  -handi  names. 

{b)  The  use  of  the  term,  nyai'  nggwail.  This  term  means 
literally  "father  and  father's  father",  but  it  is  not  used  as 
a  term  of  address  within  the  clan  except  in  invoking  the 
ancestors  collectively.  It  is,  however,  the  ordinary  term  of 
address  used  reciprocally  between  wau  and  laua.  The  term 
which  is  correspondingly  used  between  members  of  the  same 
patrilineal  group  is  bandi,  which  also  means  "novice", 
"initiate"  or  "young  man"  ^ 

(c)  The  custom  of  tshat  kundi.  " Nyai'  nggwail''  is  the 
ordinary  form  of  address  used  by  the  wau  in  talking  to  his 

^  One  informant  in  Palimbai  told  me  that  the  name  used  in  sorcery 
is  not  the  -aivan  name,  but  the  -bandi  name.  The  word  bandi  means  an 
initiate  (i.e.  either  a  novice  or  an  initiated  person)  and  names  with  this 
suffix  are  given  to  novices  by  their  wans  at  the  end  of  initiation.  The  same 
informant  said  that  women  and  children  who  have  no  -bandi  names 
could  therefore  not  be  killed  by  this  species  of  magic.  The  -bandi  name, 
like  the  -aiuan  name,  refers  to  ancestors  of  the  maternal  clan. 

^  There  is  a  curious  tangle  here,  which  I  have  not  been  able  to  unravel. 
In  the  light  of  the  general  identification  between  laua  and  ancestors  it  is 
understandable  that  the  zvau  should  call  his  laua  "father  and  father's 
father".  But  it  is  not  at  all  clear  why  the  laua  should  use  the  same  term 
for  the  loan.  Still  less  is  it  clear  why  brothers  and  fathers  should  address 
each  other  by  the  term  bandi,  seeing  that  the  -bandi  name  is  given  by  the 
wau.  The  term,  bandi,  is  definitely  more  intimate  than  nyaV  nggwail. 


44  Cultural  Premises  relevant  to 

laua,  but  on  more  marked  occasions,  when  he  is  greeting 
him,  haihng  him,  congratulating  him  or  bidding  him  farewell, 
the  wau  will  use  more  elaborate  forms  of  words.  These  are 
known  as  tshat  kundi.^  In  its  simplest  forms,  used,  e.g., 
when  a  ivau  is  saying  goodnight  to  his  laua,  tshat  kundi 
consists  simply  in  addressing  the  laua  as  a  pair  of  objects 
which  are  ancestors  of  the  wau's  clan.  The  wau  may  say — 
"You  stay,  coconut  and  areca  palm,  I  am  going  to  sleep." 
In  this  phrase  the  palms,  by  the  everyday  words  for  which 
the  laua  is  addressed,  are  totemic  ancestors  of  the  wau's 
clan. 

On  more  important  occasions  the  2vau  will  use  not  a  pair 
of  everyday  words  for  the  totemic  objects  but  a  whole  string 
of  the  personal  names  of  some  important  totemic  object  or 
ancestral  hero.  He  will  greet  his  laua  from  a  distance  as  he 
enters  the  village,  shouting  to  him : 

"Come,  Father  and  Father's  father,  Tepmeaman  and 
Kambuguli,  Weimandemi  and  Tanggulindemi,  Ulakavi  and 
Tshugukavi,  Weiuli  and  Weikama,  Tshugutshugu  and  Ngga- 
knggak,  come.  Father  and  Father's  father.  You  are  my 
laua."^ 

In  this  speech  the  names  are  those  of  the  Borassus  palm, 
which  is  also  a  fish  (cf.  p.  234)  and  an  ancestor  of  the 
Mwailambu  clan.  A  wau  who  is  a  member  of  this  clan  might 
use  this  series  of  names  in  hailing  his  laua,  but  there  are  also 
many  other  strings  of  names  of  important  ancestors  of  this 
clan,  and  any  of  these  he  might  use  for  the  same  purpose. 
He  might  even  use  several  strings  of  names  one  after  the 
other.  Members  of  other  clans  would  use  the  strings  of  names 
of  their  special  ancestors. 

Most  of  the  names  which  are  used  in  tshat  kundi  might 
also  be  given  to  sons  of  the  wau's  clan.  But  there  are  also 
a  few  strings  of  names  which  have  apparently  dropped  out 

^  Tshat  is  a  verb  which  means  to  "step  across"  something,  as,  e.g.  in 
the  naven  ceremonial,  the  hero  "steps  across"  the  women.  Kundi  is  the 
ordinary  word  for  " mouth ",  "speech ",  "language ",  etc.  There  is  another 
phrase  for  the  same  custom — tshivera  kundi — but  I  do  not  know  the  literal 
translation  of  this. 

^  In  this  speech,  I  have  inserted  the  word  "and"  between  the  names 
of  each  pair.  In  the  declaiming  of  the  speech,  the  intonation  would 
give  the  pairing  and  no  copula  would  be  used. 


the  Wau-laua  Relationship  45 

of  use  in  the  naming  of  children  and  are  now  only  used  for 
tshat  kundi. 

Sometimes  tshat  kundi  is  used  by  persons  other  than  the 
wau,  e.g.  as  a  device  for  emphasising  the  fact  that  the  person 
addressed  is  a  "brother"  of  the  speaker  in  the  sense  that 
both  are  lauas,  of  the  same  clan.  In  all  cases  the  names  used 
are  those  of  ancestors  of  the  maternal  clan  of  the  person 
addressed. 

Lastly,  the  names  used  in  tshat  kundi  are  chosen  merely- 
because  they  refer  to  important  ancestors  of  the  rriaternal 
clan.  The  choice  is  not  guided  in  any  way  by  the  particular 
-awan  or  -bandi  names  which  have  been  given  to  the  person 
addressed — except  in  so  far  as  the  -awan  name  indicates  to 
what  clan  the  person's  mother  (or  classificatory  mother) 
belongs. 

Thus  we  may  sum  up  the  custom  of  tshat  kundi  by  saying 
that  the  laua  is  ceremonially  addressed  as  if  he  were  a  random 
collection  of  his  mother's  important  ancestors. 

[d)  The  laua  dances  in  masks  which  represent  the  ancestors 
of  the  maternal  clan.  He  blows  the  flutes  which  are  their 
ancestors.  He  carves  representations  of  their  ancestors  on  the 
house-posts  which  are  to  stand  in  their  ceremonial  house. 
He  carves  the  ancestral  mwai  heads  (cf.  Plate  XXVHI  b)  which 
are  to  adorn  the  masks  of  the  maternal  clan — the  masks  in 
which  he  will  later  dance. 

{e)  The  laua  is  entitled  to  ornament  his  body  with  those 
plants  which  are  totemic  ancestors  of  his  maternal  clan ;  and 
he  will  resent  the  use  of  these  plants  by  persons  not  so  entitled. 

(/)  When  the  members  of  a  clan  sacrifice  to  their  clan 
ancestors,  it  is  the  lauas  who  eat  the  sacrifice. 

A  similar  position  is  brought  out  with  diagrammatic  clear- 
ness in  the  procedure  in  cases  of  offence  against  the  rights  of 
land  tenure. 

The  land  of  a  clan  is  watched  over  by  the  angk-au.  This 
word  means  literally  "potsherds"  and  is  a  term  for  those 
spirits  of  ancestors  which  are  symbolised  by  the  old  broken 
sherds  which  lie  under  the  house,  sherds  which  have  been 
thrown  away  by  generations  of  previous  inhabitants  of  the 
house.  When  the  angk-au  see  any  outside  person  violating 


46  Cultural  Premises  relevant  to 

the  land  of  the  clan,  e.g.  planting  upon  it  or  stealing  its  pro- 
ducts, they  make  a  little  noise  in  the  house  at  night,  "like 
a  rat  squeaking".  The  owner  of  the  house  hears  this  sound 
and  goes  out  next  day  to  see  what  is  amiss.  He  finds  that 
somebody  has  planted  coconuts  on  his  land.  Then  he  puts 
ndjambwia  on  the  ground.  Ndjamhzoia  may  be  spikes  set  up 
to  wound  the  foot  of  the  trespasser,  or  they  may  be  magical 
objects  to  cause  sickness  of  the  trespasser;  the  same  term  is 
used  for  both  techniques.  Later  the  trespasser  comes  and  his 
feet  are  wounded  and  develop  sores.  The  angk-au  will  pre- 
vent the  healing  of  these  sores  till  the  offender  goes  to  the 
members  of  the  clan  who  own  the  land.  Then  the  offender 
will  kill  a  fowl  and  present  coconuts  to  them.  He  will  also 
go  to  the  patch  of  ground  and  will  there  hang  up  a  basket  of 
valuables  and  a  tamhointsha}  The  angk-au  will  take  the 
"soul"  of  the  valuables  and  of  the  tambointsha;  and  after 
a  few  days  the  valuables  will  be  returned  to  the  trespasser. 
But  the  fowl  and  coconuts  will  be  eaten  by  the  lauas  of  the 
clan  which  owned  the  ground. 

(g)  The  laua  eats  his  maternal  ancestors.  On  one  of  the 
lakes  in  the  neighbourhood  of  Mindimbit  village  there  is  an 
island  of  floating  grass  or  sud  (agwi) ;  and  on  this  a  certain 
crocodile  lays  its  eggs.  This  crocodile  has  the  personal  name, 
Mwandi-ntshin,  and  is  an  ancestor  of  the  Mandali  clan. 
Members  of  this  clan  go  from  the  village  to  collect  the  eggs. 
It  is  said  that  they  find  as  many  as  sixty  eggs  and  among  them 
two  double  eggs.  The  ordinary  eggs  they  eat  but  the  double 
ones  are  presented  to  their  lauas,  who  eat  them. 

In  this  detail  we  have  an  isolated  case  of  the  laua  eating 
wau's  ancestors.  But  besides  this  there  is  also  a  regular 
ritual  in  which  the  same  act  is  performed.  This  ritual  is 
called  pwivu.  The  maternal  clan  prepare  in  a  large  bowl  a 
mixture  containing  scrapings  from  the  bones  of  their  ancestors 

^  Tambointsha  are  tassels  of  feathers  tied  to  string.  They  are  symbols 
of  successful  homicide  and  worn  on  the  lime  stick  as  a  tally  of  the  owner's 
successful  kills.  A  kill  may  be  scored  for  homicide  achieved  by  means  of 
ndja?nbwia,  but  not  for  homicides  achieved  by  other  magical  techniques. 
The  act  of  presenting  a  tambointsha  to  the  angk-au  thus  falls  into  the 
same  pattern  as  the  presentation  of  tambointsha  described  on  p.  99. 
In  both  cases  we  may  see  a  formal  recognition  of  the  fact  that  the  damage 
done  was  a  legitimate  act  of  vengeance. 


the  Wau-laua  Relationship  47 

and  other  curious  clan  relics.  This  mixture  is  eaten  by  the 
children  of  the  village — but  especially  by  the  lauas  of  the 
clan  whose  ancestors  have  gone  into  the  mixture.^ 

{h)  Finally,  when  the  laua  dies  a  ceremony  called  min- 
tshanggu  is  performed.  A  figure,  which  has  for  head  the 
modelled  and  decorated  skull  of  the  laua,  is  set  up  on  a 
platform  which  is  suspended  from  the  roof  of  the  house  so 
that  at  a  push  it  will  swing  to  and  fro.  This  platform  is  called 
the  agzvi  (floating  grass  island)  and  the  spectacle  represents 
the  voyage  of  the  ghost,  on  a  drifting  patch  of  floating  grass, 
down  the  Sepik  River  to  the  land  of  the  dead.  The  platform 
is  suspended  about  two  feet  from  the  floor  and  from  its  edge 
hangs  down  a  fringe  of  strips  of  palm  leaves,  forming  a  screen 
so  that  the  women  in  the  house  cannot  see  into  the  space 
under  the  platform.  This  fringe  is  called  tshimbwora,  the 
word  which  is  used  for  water  in  the  special  ritual  jargon  of 
the  shamanic  spirits. 

The  ritual  which  accompanies  this  spectacle  is  divided 
into  two  halves,  the  first  half  being  performed  by  the  members 
of  the  dead  man's  clan,  and  the  second  half  by  members  of 
his  mother's  clan.  The  two  halves  are  essentially  similar,  each 
consisting  in  the  singing  of  name  songs.  In  the  particular 
celebration  which  I  witnessed,  the  members  of  the  dead 
man's  clan  collected  together  in  front  of  the  agwi  soon  after 
dusk  and  sang  name  songs  till  after  dawn  next  morning.  Their 
clan  flutes  had  been  smuggled  into  the  house  and  hidden 
under  the  agwi.  From  time  to  time  the  succession  of  name 
songs  was  varied  by  music  from  their  flutes  played  by  men 
(their  lauas)  hidden  under  the  agwi.  The  same  hidden  men 
also  occasionally  caused  the  suspended  agwi  to  swing,  repre- 
senting the  rocking  caused  by  waves  in  the  water. 

Early  next  morning  the  members  of  the  dead  man's  clan 
dispersed  and  in  their  place  the  members  of  his  mother's  clan 
assembled.  They  brought  with  them  a  number  of  totemic 

^  A  more  detailed  account  of  the  pwivu  ritual  is  published  in  Oceania, 
1932,  p.  472.  Among  the  latmul  the  eating  of  the  maternal  clan  is 
only  found  in  these  somewhat  symbolic  forms.  But  latmul  informants 
told  me  that  among  a  neighbouring  tribe,  on  the  Sud  River,  it  is  usual 
for  the  sister's  child  to  eat  a  piece  of  flesh  from  the  thigh  of  his  mother's 
dead  brother. 


48  Cultural  Premises  relevant  to 

objects,  their  ancestors — branches  of  trees,  paddles,  etc. 
These  they  added  to  the  decorations  of  the  agwi,  which  had 
already  been  decorated  with  the  totemic  emblems  of  the 
dead  man's  clan.  They  also  brought  a  hooked  stick  and  their 
own  clan  flutes  which  were  substituted  for  the  others  under 
the  agzci.  Then  they  began  to  sing  their  name  songs  and  con- 
tinued in  this  until  late  afternoon.  During  the  singing,  the 
hooked  stick  was  used  from  time  to  time  to  swing  the  agwi: 
the  members  of  the  maternal  clan  overtly  pulling  the  agwi 
and  the  figure  of  the  dead  towards  themselves. 

The  three  premises  which  I  have  mentioned  deal  with 
the  relationships  of  the  child  to  its  father,  mother  and 
maternal  clan.  These  may  be  summed  up  in  a  form  especially 
relevant  to  Jiaven  by  saying  that  the  child  is  closely  identified 
with  its  father  but  competes  with  him.  The  child's  identity 
with  its  mother  and  its  link  with  the  maternal  clan  are  more 
obscure.  But  here  the  child  is  not  a  competitor  but,  rather, 
an  achievement. of  the  mother;  and  the  child's  achievements 
are  her  achievements,  the  triumphs  of  her  clan. 

This  conclusion  can  be  further  documented  from  another 
mortuary  ritual.  When  a  great  man  dies  a  figure  is  set  up  by 
the  members  of  his  initiatory  moiety  to  represent  him  and  is 
decorated  with  symbols  of  all  his  achievements.  Spears  are 
set  up  to  the  number  of  his  kills  and  baskets  are  suspended 
from  the  shoulder  of  the  figure  to  the  number  of  his  wives. 
But  no  symbols  are  added  to  show  how  many  children  he  has 
begotten.  For  a  woman,  when  she  dies,  a  post  [nggambut) 
is  set  up  and  is  decorated  with  the  ornaments  which  were 
given  to  her  sons  by  the  senior  age  grade  at  initiation.  Her 
greatness  is  vicarious  and  lies  in  the  achievements  of  her 
sons. 

Similarly  in  the  list  of  occasions  which  are  celebrated  by 
naven,  we  find  that  naven  are  celebrated  for  a  boy  when  he 
takes  a  wife,  but  not  when  his  first  son  is  born.  For  a  girl 
naven  are  celebrated  when  she  gives  birth  to  a  child,  but  not 
for  her  marriage. 

There  is  a  curious  myth  which  is  probably  relevant  in  this 
connection,  as  showing  what  might  happen  if  the  naven 
patterns  were  disregarded.   In  the  myth  naven  are  celebrated 


the  Wau-laua  Relationship  49 

for  a  man  by  his  wife's  brother  and  wife's  father  on  the 
occasions  of  the  birth  of  both  his  first  and  second  children. 
These  relatives  present  valuables  to  the  husband  and  rub 
their  buttocks  on  his  shin.  He  in  return  presents  food  to 
them  (i.e.  the  normal  pattern  of  presentations  is  reversed). 
After  the  second  naven  the  relatives  stay  and  stay  demanding 
more  food,  until  the  husband  says,  "There  is  none  left." 
Then  they  say,  "You  are  lying."  The  husband  said,  "I  am 
not  lying  ",  and  when  he  had  said  that,  he  would  have  speared 
the  child.    But  his  wife — it  was  she  who  speared  the  child. 

With  this  clue  to  the  general  sense  of  these  formulations, 
the  procedure  at  the  little  boy's  first  kill  falls  into  place  as 
consistent  with  the  general  pattern.  The  boy's  triumph  is 
the  triumph  of  his  maternal  clan;  therefore  it  is  allowable 
for  his  wau  to  help  him  in  the  holding  of  the  spear.  But  if 
this  office  were  performed  by  the  father  we  may  suppose  that 
the  kill  would  be  claimed  and  celebrated  by  the  father's 
maternal  clan. 

Lastly,  we  may  enquire  into  an  apparent  discrepancy  in 
the  premises  which  pattern  the  relationship  of  a  child  to  its 
mother's  clan.  On  the  one  hand  we  have  details  of  behaviour 
which  indicate  that  the  child  is  the  achievement  of  the 
mother  and  its  achievements  are  her  achievements,  and  on 
the  other  hand  we  have  details  which  indicate  that  the  child 
is  the  ancestors  of  its  mother's  clan.  This,  I  believe,  is  a 
problem  which  cannot  be  satisfactorily  solved  until  we  have 
examined  the  ethos  of  latmul  culture.  In  later  chapters,  the 
association  between  pride  and  ancestors  will  be  documented ; 
and  I  will  here  overstep  the  strict  limits  of  structural  enquiry 
by  suggesting  that  pride  is  the  factor  which  links  together 
achievement  and  ancestors,  two  apparently  different  ideas, 
alike  only  in  their  emotional  content. 

4.  That  there  is  an  element  of  identity  between  a  brother 
and  his  sister.  This  premise  like  the  last  is  shown  dia- 
grammatically  in  the  system  of  personal  names.  Siblings  of 
opposite  sex  are  frequently  given  the  same  name  with  dif- 
ferences only  in  gender  termination  (cf.  p.  243).  The  same 
formulation  is  recognisable  in  the  formation  of  a  compound 
kinship  term  of  the  type  to  which  I  have  already  referred. 

4 


50  Cultural  Premises  relevant  to 

This  is  the  term  wau-nyame  (mother's  brother  and  mother), 
which  is  used  as  a  collective  term  for  the  maternal  clan,  as  a 
synonym  in  fact  of  the  term  wau-mhuambo  to  which  I  referred 
above.  The  same  identification  is  also  indicated  by  the  terms 
used  between  brothers-in-law.  These  terms  are  lan-ndo  for 
sister's  husband,  and  tawontu  (i.e.  tagioa-ndo)  for  wife's 
brother.  The  term  Ian  means  husband  and  the  term  tagwa 
means  wife.  Thus  the  sister's  husband  is  called  "husband- 
man", the  male  speaker  identifying  himself  with  his  own 
sister  in  their  relationship  to  her  husband ;  while  reciprocally, 
the  wife's  brother  is  called  "wife-man",  the  speaker  identi- 
fying wife's  brother  with  wife.  Both  these  terms  therefore 
imply  the  identification  of  the  brother  and  sister.  A  concrete 
example  of  the  working  of  this  identification  may  be  seen  in 
the  myth  of  the  brother  and  sister  who  went  together  to  work 
sago,  quoted  above. 

In  this  identification  of  siblings  of  opposite  sex  there  is  a 
type  of  insistence  of  which  we  shall  meet  other  cases  later, 
an  insistence  that  the  mere  genealogical  position  is  not 
enough ;  the  identification  must  be  made  real  in  behaviour — 
more  especially  in  ceremonial  behaviour.  In  the  passing  on  of 
clan  names,  which  are  given  on  the  basis  of  the  patrilineal 
alternation  of  generations,  a  man  has  a  right  to  give  the  names 
of  his  classificatory  and  own  fathers  to  his  sons.  But  the 
women  of  the  clan  also  have  clan  names  and  a  man  cannot 
simply  take  his  father's  deceased  sister's  names  and  give  them 
to  his  daughters.^  Before  this  can  happen  the  identification 
of  a  brother  with  the  woman  must  apparently  be  demon- 
strated in  ceremonial  behaviour.  This  is  done  either  im- 
mediately after  the  woman's  death  or  after  the  death  of  her 
husband  or  her  brother : 

^  This  would  seem  to  be  the  general  sense  of  the  statements  which 
I  collected  about  this  matter:  one  informant,  however,  stated  that  not 
only  the  female  names  but  also  the  male  names  must  be  got  from  the 
father's  sister.  The  case  which  we  were  discussing  was  one  in  which  a 
man  on  the  day  after  his  father's  death  went  and  cut  his  father's  sister's 
hair,  "buying  both  his  father's  and  her  name"  with  a  Turbo  shell. 

I  do  not  know  the  meaning  of  the  symbolic  water  into  which  the  Turbo 
shell  is  put;  it  may  perhaps  be  a  reference  to  the  mythological  origin  of 
these  shells  from  the  waves  of  the  sea  and  a  genital  symbol.  This  piece 
of  ceremonial  is  the  only  context  I  know  of  in  which  valuables  are 
ceremonially  given  to  members  of  own  clan. 


the  Wau-laua  Relationship  51 

{a)  When  a  woman  dies,  her  brother  will  go  to  the  corpse 
and  will  dress  her  with  a  skirt,  paint  her  face,  and  attend  to 
the  details  of  her  funeral.  He  will  then  say,  "  I  am  the  man 
who  will  call  the  names." 

{b)  When  a  man  dies,  his  son  may  go  to  the  dead  man's 
sister  in  her  house  shortly  after  the  death.  He  puts  a  shell 
valuable,  generally  a  Turbo  shell,  into  water  in  one  of  her 
pots  and  he  cuts  her  hair  for  her.  The  shell  becomes  her 
property  and  the  brother's  son  gains  the  right  to  give  the 
names  to  his  children.  (In  the  actual  case  in  which  this 
procedure  was  followed,  the  woman  had  no  other  surviving 
brothers.) 

(c)  When  a  man  dies,  the  widow's  hair  will  be  cut  and  a 
shell  valuable  presented  to  her  in  water  by  either  her  brother 
or  her  brother's  son.  (In  the  texts  in  which  this  procedure 
is  mentioned  there  is,  however,  no  mention  of  the  right  to 
the  names  and  it  is  possible  that  in  this  context  the  cere- 
monial is  merely  an  affirmation  of  the  widow's  membership 
of  the  clan  into  which  she  was  born  but  from  which  she  has 
been  to  some  extent  separated  by  marriage.) 

{d)  If  a  woman  is  left  the  solitary  survivor  of  her  clan, 
all  the  names  of  that  clan  become  vested  in  her  and  her  bride 
price  becomes  correspondingly  great  since  the  right  to  give 
the  names  will  fall  to  her  husband  or  her  children.  In  such 
cases,  an  effort  is  made  by  the  members  of  closely  related 
clans,  her  distant  classificatory  brothers,  to  get  her  as  wife 
for  one  of  them,  so  as  to  retain  the  names  within  the  phratry. 
In  this  procedure  we  may  see  weak  identification,  based  only 
upon  classificatory  brotherhood,  being  strengthened  by  the 
additional  identification  based  upon  marriage. ^ 

^  As  might  be  expected  from  the  complexity  of  these  observances, 
quarrels  sometimes  arise  over  the  matter  of  names  held  by  women. 
I  came  across  one  dispute  of  this  sort:  The  two  last  male  members  of 
the  Mbe  clan  had  been  killed  by  a  raiding  party  some  years  before,  leaving 
only  two  women.  An  unrelated  man,  Tshamelwan,  had  provided  the 
food  (?  or  valuables)  for  the  mortuary  feast  for  these  men  and  on  the 
strength  of  this  he  had  claimed — "  I  take  the  names ! "  and  had  proceeded 
to  give  Mbe  names  to  his  sons. 

But  there  still  remained  the  two  women,  by  name  Teli  and  Tampiam. 
Teli  was  an  old  widow,  with  two  sons  to  whom  she  had  given  some 
Mbe  names.  Tampiam  was  a  young  woman. 

The   clan,    lavo,   closely   linked   with   Mbe,   therefore   arranged   that 


52  Cultural  Premises  relevant  to 

5.  That  a  woman  is  to  some  extent  identified  with  her 
husband.  In  this  case  we  have  to  consider  an  identification 
which  is  not  reciprocal.  It  seems  that  the  wife  does  take  over 
a  part  of  her  husband's  status,  but  it  seems  that  he  takes 
over  little,  if  any,  of  hers :  an  asymmetry  which  is  consistent 
with  preponderant  patriliny  and  patrilocal  marriage.  In  our 
analysis  of  the  naven  ceremonial  we  shall,  however,  be  con- 
cerned only  with  cases  of  wives  taking  status  from  their 
husbands ;  and  that  only  a  propos  oimbora  (mother's  brother's 
wife)  and  tshaishi  (elder  brother's  wife).  This  formulation 
is  not  relevant  to  the  patterning  of  wau's  behaviour  in  naven. 

The  identification  of  wife  with  husband  is  in  general  very 
little  summarised  in  the  culture.  We  do  not  find,  for  example, 
any  tricks  of  the  naming  system  in  which  this  identification 
is  expressed  or  assumed.  Neither  partner  changes  his  or  her 
name  at  marriage;  and,  though  new  kinship  terms  may  be 
applied  to  the  partners  as  a  result  of  the  marriage,  these  new 
terms  for  the  most  part  still  discriminate  between  the  partners. 
This  is  true  of  all  the  marriages  of  closer  relatives,  including 
all  marriages  of  siblings  of  the  parents  and  of  persons  who  are 
identified  with  these  in  classificatory  terminology.  The  sole 
exception  to  this  among  the  near  relatives  is  the  use  of  the 
term  naisagut  for  wife's  father  and  his  wives. 

But  in  the  less  important  ramifications  of  kinship  we  find 
that  the  wife  is  referred  to  by  the  same  term  as  the  husband, 
e.g.  mbuambo  is  the  term  used  for  mother's  father  and  his 
wives,  as  well  as  for  mother's  brother's  son  and  his  wives. 
Similarly  kaishe-ndo  and  kaishe-ragzva,  terms  differing  only 
in  gender  termination,  are  used  reciprocally  between  the 
parents  of  a  husband  and  the  parents  of  his  wife. 

In  the  case  of  tshaishi  (elder  brother's  wife)  the  terminology 
is  interesting.   She  always  identifies  herself  with  her  husband 

Tampiam  should  marry  Tshava  (my  informant),  a  young  man  of  lavo, 
so  as  to  keep  the  Mbe  names  in  the  phratry. 

No  children  had  been  born  of  this  marriage  when  I  was  in  Kankanamun 
and  Tampiam  was  still  angry  with  Tshamelwan  for  stealing  the  Mbe 
names.  Lacking  children,  she  was  doing  her  best  to  reserve  the  names 
by  giving  them  to  her  pigs,  canoe,  etc.  She  also  attributed  her  sterility 
to  sorcery  and  suspected  Tshamelwan  of  being  the  cause  of  it. 

Tshava  stated  that  Tshamelwan  repudiated  this  charge,  suggesting  that 
the  sterility  was  caused  by  Teli. 


the  Wau-laua  Relationship  53 

in  calling  his  younger  brother  tshuambo,  a  term  which  is 
otherwise  reserved  for  younger  siblings  of  the  same  sex  as 
the  speaker.  The  ordinary  reciprocal  to  tshuambo  is  nyamun 
(elder  sibling  of  same  sex),  but  the  husband's  younger  brother 
does  not  use  this  term  for  his  tshaishi,  except  in  the  naven 
ceremonial. 

The  identification  of  man  and  wife  is  somewhat  more 
evident  in  everyday  life  than  in  the  terminology  of  kinship. 
As  we  have  seen,  husband  and  wife  are  discriminated  in  their 
contacts  with  near  kin  but  identified  when  they  are  dealing 
with  outsiders.  The  picture  of  everyday  life  fits  with  this, 
inasmuch  as  the  contexts  in  which  we  see  the  couple 
identified  are  especially  those  in  which  they  are  dealing  with 
outside  persons  or  inanimate  objects.  We  find  them,  for 
example,  working  together,  sharing  a  house ^  and  acting  as 
each  other's  agents  in  barter.  There  is  a  very  close  mutual 
economic  dependence  between  man  and  wife;  and  though 
every  piece  of  property  can  be  defined  as  belonging  either 
to  him  or  to  her,  it  is  almost  possible  to  see  the  household 
as  a  single  economic  unit.  This  may  be  illustrated  by  the 
case  of  a  man  in  Komindimbit  village  who  had  lost  all  his 
own  and  his  wife's  property  in  a  fire  which  had  destroyed 
his  house.  I  asked  him  how  he  would  reinstate  himself  and 
he  replied  immediately,  "My  wife  will  feed  up  pigs  and  I 
shall  sell  them".  Similarly,  if  a  wife  comes  without  dowry 
to  the  household,  her  co-wives  will  hold  it  against  her,  as  a 
reproach,  that  she  contributed  nothing  to  the  common  wealth 
of  the  group. ^ 

^  It  may  be  noted  that  the  husband  is  often  absent  from  the  house  and 
spends  much  of  his  time  in  the  ceremonial  house  with  the  other  men. 
It  is  not  approved  of  in  a  man  that  he  should  spend  a  great  part  of  the 
day  in  his  dwelling  house  among  the  womenfolk. 

-  At  this  point  we  may  note  a  weakness  in  the  argument.  We  shall  later 
regard  the  behaviour  of  mbora  and  tshaishi  as  consistent  with  this  formula- 
tion. But  I  am  not  able  to  document  the  identification  of  man  and  wife 
in  those  contexts  in  which  they  are  dealing  with  the  husband's  younger 
brother  and  the  husband's  sister's  son.  JVIan  and  wife  are  identified 
when  they  are  dealing  with  outside  people  but  it  is  not  clear  that  they 
are  identified  in  those  contexts  which  are  specially  relevant  to  the 
present  enquiry.  The  concept  of  context  must  certainly  be  added  to  that 
of  identification  and  we  should  be  chary  of  using  the  latter  term  without 
some  indication  of  the  contexts  in  which  the  persons  are  identified. 


CHAPTER  V 

Sorcery  and  Vengeance 

IN  the  previous  chapter  I  have  stated  that  in  the  latmul 
kinship  system  there  are  four  identifications  which  we 
may  regard  as  major  premises  of  the  system : 

1.  An  identification  between  father  and  son. 

2.  An  identification  (on  different  Hues)  between  a  boy  and 
his  mother's  clan. 

3.  An  identification  between  brother  and  sister. 

4.  An  identification  between  husband  and  wife. 

The  working  of  these  premises  will  be  illustrated  in  the 
present  chapter  from  material  which  I  collected  about  a 
special  context  of  latmul  culture,  that  of  retribution. 

In  the  investigation  of  latmul  ideas  about  sorcery  and 
retribution,  we  at  once  come  upon  the  native  concept, 
ngglambi,  a  word  which  we  might  translate  into  abstract 
terms  as  "dangerous  and  infectious  guilt".  But  in  native 
thought  ngglambi  is  much  more  concrete  than  this.  It  is 
thought  of  as  a  dark  cloud  which  envelops  a  man's  house 
when  he  has  committed  some  outrage.  This  cloud  can  be 
seen  by  certain  specialists  who,  when  they  are  consulted  as 
to  the  cause  of  some  illness  or  disaster,  rub  their  eyes  with 
the  white  undersides  of  the  leaves  of  a  tree,  and  are  then 
able  to  see  the  dark  cloud  hovering  over  the  house  of  the 
person  whose  guilt  is  responsible  for  the  sickness.  Other 
specialists  are  able  to  smell  ngglambi  and  say  that  it  has  a 
"smell  of  death — like  a  dead  snake".  It  is  characteristic  of 
ngglambi  that  it  may  "turn  and  go",  causing  the  sickness 
either  of  the  guilty  person  or  of  his  relatives.  It  is  this  trans- 
mission of  ngglambi  which  is  relevant  to  the  questions  of 
identification  between  kin. 

As  to  the  agency  responsible  for  the  dangers  of  ngglambi, 
the  people  are  very  articulate  and  their  statements  fall  into 
four  closely  related  groups. ^ 

^  There  is  a  fifth  phrasing  of  the  matter  which,  though  irrelevant  to 
the  vahdity  of  ngglambi  infection  as  an  index  of  identification,  is  yet 
interesting  for  the  Hght  it  throws  upon  latmul  sorcery.     It  is  believed 


Sorcery  and  Vengeance  55 

{a)  It  is  said  that  the  ngglamhi  itself  causes  illness  and 
death.  This  is  an  absolutely  impersonal  theory  of  the  agentless 
causation  of  disease  and  is  tantamount  to  the  statement, 
which  the  latmul  also  make :  that  a  man's  sorcery  will  of  itself 
return  to  plague  him.  This  type  of  phrasing  may  almost  be 
regarded  as  a  euphemism  for  the  more  personal  statements 
which  follow. 

{b)  It  is  said  that  the  illness  is  caused  by  certain  spirits, 
wagan,  who  avenge  various  offences,  acting  upon  their  own 
volition  but  generally  in  the  interests  of  their  clan  descen- 
dants. 

(c)  It  is  said  that  the  persons  whom  the  guilty  man  has 
offended  will  complain  to  their  clan  zvagan,  who  will  avenge 
the  offence.  Such  complaints  may  take  various  forms:  in 
some  cases  it  is  apparently  sufficient  if  the  zvagan  hears  the 
weeping  of  his  descendants;  when  he  hears  he  will  at  once 
set  about  avenging  the  injury  which  has  been  done  to  them. 
Failing  this,  the  offended  people  will  beat  a  rhythm  on  the 
nggambut  (mortuary  post)  of  the  person  whom  they  wish 
avenged.  This  rhythm  would  be  the  slit-gong  call  (tavet)  for 
summoning  the  wagan.  The  latter  would  hear  the  rhythm 
"and  having  heard  it  he  would  not  forget.  There  would  be 
no  talk  about  the  matter.  To-morrow  the  wagan  would  come 
to  smell  out  the  ngglamhi" .  Alternatively  the  offended  people 
might  go  to  the  wagan  personally  at  a  time  when  he  possesses 
his  shaman.  They  would  talk  to  the  possessed  man  (i.e.  to 
the  wagan)  and  pay  him  areca  nuts  and  shell  valuables  for 
his  services  in  avenging  the  offence. 

{d)  It  is  said  that  the  offended  persons  will  themselves 
practise  sorcery  which  will  cause  the  sickness  or  death  of  the 
guilty  person  or  of  his  relative.  Alternatively  they  may  pay 
a  professional  sorcerer  to  do  this  with  or  without  the  aid 
of  his  wagan. ^ 

that  a  cunning  sorcerer  will  kill  off  victims  in  pairs.  One  of  the  killed  will 
be  the  person  against  whom  the  sorcerer  has  a  grudge,  while  for  the 
other  he  will  choose  one  of  his  own  relatives — thus  "short-circuiting" 
the  retribution  and  reserving  to  himself  some  choice  as  to  which  of  his 
relatives  shall  die  to  pay  for  the  desired  death  of  his  enemy. 

^  In  this  community  where  violent  death  is  frequent,  death  by  sickness 
(i.e.  sorcery)  is  not  very  seriously  regarded.  There  is  very  little  of  the 
almost  paranoid  fear  of  sorcery  which  is  characteristic  of  many  primitive 
cultures. 


56  Sorcery  and  Vengeance 

In  all  these  cases  the  resulting  sickness  may  be  described 
as  due  to  the  ?igglambi  of  the  guilty  man.  Thus  the  theory 
of  ngglambi  fluctuates  between  a  belief  in  the  spontaneous 
effects  of  guilt  which  may  be  "infectious"  and  concrete 
statements  that  so-and-so  will  avenge  himself  either  upon  the 
offender  or  upon  his  relatives. 

In  the  light  of  this  double  meaning  of  ngglambi  it  is  inter- 
esting to  see  which  of  a  guilty  man's  relatives  are  most  likely 
to  be  imperilled  by  this  dark  cloud — since  we  may  reasonably 
translate  a  statement,  that  .^'s  ngglambi  has  caused  the  sick- 
ness of  B,  into  our  jargon  of  identification:  ^'  A  has  offended 
against  C ;  C  identifies  B  with  A  and  proceeds  to  avenge  him- 
self upon  A  by  practising  sorcery  upon  5." 

But  though  this  would  seem  to  be  the  natural  interpreta- 
tion of  ngglambi  infection,  the  theories  of  retribution  contain 
another  emphasis  which  would,  at  first  sight,  seem  to  confuse 
the  issue.  This  is  the  emphasis  upon  the  law  of  "an  eye  for 
an  eye  and  a  tooth  for  a  tooth".  For  example,  if  A  magically 
kills  JB's  wife,  B  will  be  angry  and  will  avenge  himself  by 
killing  yi's  wife.  Such  an  exact  retribution  the  latmul  would 
approve  of  ;^  but  they  would  still  phrase  the  death  of  ^'s 
wife  as  due  to  ^'s  ngglambi,  although  in  killing  the  woman 
B  was  not  identifying  her  with  her  husband.  But  such  a  con- 
fusion of  the  issue  is,  I  think,  more  apparent  than  real.  The 
death  of  A's  wife  is  still  due  to  an  identification  of  a  husband 
with  his  wife :  B  identified  himself  with  his  wife  sufficiently 
to  avenge  her  death,  and  the  lex  talionis  indicated  that  the 
correct  person  for  him  to  kill  was  ^'s  wife.    In  place  of  the 

^  In  such  a  case,  the  avenger  will  first  perform  the  magic  in  secret, 
but  when  the  woman  is  dead  he  will  stand  up  in  the  ceremonial  house  and 
proclaim  what  he  has  done  and  his  justification  for  doing  it.  If  the 
lex  talionis  has  been  accurately  followed,  this  may  be  the  end  of  the 
matter ;  otherwise  a  long  sorcery  feud  may  result  accounting  for  as  many 
as  fifteen  deaths. 

As  regards  the  effectiveness  of  sorcery  and  ngglambi  in  causing  sickness 
and  death,  I  have  adopted  the  native  phrasing  of  the  matter.  From  a 
European  point  of  view  it  is  probable  that  most  of  the  deaths  in  a  latmul 
sorcery  feud  are  due  to  "  natural  causes  "  and  the  feud  itself  may  be  almost 
fictitious.  But  in  native  belief,  the  deaths  are  ascribed  to  the  progress 
of  the  feud ;  and  the  fact  that  the  deaths  do  not  usually  follow  the  lex 
talionis  with  any  accuracy  is  ascribed  to  the  unreasonableness  of  wagan 
and  sorcerers. 


Sorcery  and  Vengeance  57 

question  "  Who  is  likely  to  be  identified  with  the  guilty  man? " 
we  have  only  substituted  the  question  "Who  is  likely  to 
identify  himself  with  the  killed  person?"  The  phrasing  of 
retribution  in  terms  of  lex  talionis  when  analysed  in  this  way 
justifies  us  in  regarding  the  ngglamhi  infection  which  occurs 
between  two  relatives  as  an  indication  that  persons  related 
in  the  same  manner  are  to  some  extent  identified. 

An  additional  piece  of  evidence  which  seems  to  corroborate 
the  theory  that  ngglamhi  and  lex  talionis  may  be  equated  from 
a  structural  point  of  view  is  afforded  by  the  homonymous 
meanings  of  the  word  tiggambzca.  This  word  means  on  the 
one  hand  "vengeance",  in  such  phrases  as  nggatnbwa  kela, 
"to  take  vengeance" — the  duty  of  the  man  whose  relative 
has  been  killed  by  enemies.  But  on  the  other  hand  the  same 
word  is  also  applied  to  certain  spirits,  kerega  nggambwa, 
literally  "eating  vengeances",  who  after  death  tear  the  flesh 
from  the  corpse  with  their  long  finger-nails.  Thus  in  the 
word  nggambwa,  as  in  the  word  ngglambi,  we  find  a  super- 
natural concept  mixed  up  with  concrete  ideas  of  vengeance. 

With  this  preliminary  examination  of  the  native  ideas 
about  sorcery  and  vengeance,  we  may  now  proceed  to  examine 
a  series  of  stories  of  sickness  and  death,  confident  that  when 
we  find  one  individual  suffering  for  the  deeds  of  another 
we  are  entitled  to  see  in  that  fact  an  instance  of  identification 
either  between  the  guilty  person  and  the  sufferer  or  between 
two  persons  related  in  the  same  way  as  these. 

The  material  here  presented  is,  strictly  speaking,  only 
relevant  in  its  present  context  for  the  light  which  it  throws 
upon  the  various  identifications,  but  inasmuch  as  the  stories 
are  also  interesting  from  other  points  of  view  I  shall  give  them 
in  rather  more  detail  than  would  be  necessary  merely  to 
document  the  identifications. 

Case  I  is  a  translation  of  a  dictated  text;  but  the  other 
stories  are  reconstructed  out  of  very  full  notes  taken  during 
slow  narration  of  the  stories  in  pidgin  English  or  in  the  native 
language  or  in  a  mixture  of  the  two.  I  have  not  scrupled  to 
insert  conjunctions  and  binding  clauses  in  these  texts  so  as 
to  make  them  more  intelligible,  but  wherever  a  more  im- 
portant explanatory  statement  has  been  inserted,  it  is  put  in 


58  Sorcery  and  Vengeance 

parentheses.    Names  of  the  people  concerned  are  given  in 
somewhat  shortened  form. 

Case  I.  This  is  an  extract  from  a  general  account  of  the 
proceedings  which  would  accompany  the  sickness  and  death 
of  a  man.  The  account  was  given  me  by  Mbana  in  Kan- 
kanamun.  It  describes  the  procedure  of  divination  for  the 
cause  of  the  sickness : 

His  (the  sick  man's)  son  and  wife  take  ashes  from  the  fire 
basin  and  set  them  in  little  heaps  on  the  spathe  of  a  palm. 
There  will  be  several  heaps  on  the  one  spathe  (one  heap  for 
each  relative  whose  offences  may  be  the  cause  of  the  sickness). 
Then  they  say,  "What  is  the  root  of  the  matter?  Why  did 
he  die?  Was  it  because  his  father  went  to  other  people's 
wives?  Was  it  because  of  the  doings  of  his  father's  father 
that  he  died?" 

Then  the  palm  spathe  with  the  heaps  of  ashes  is  put  into 
a  mosquito  bag.  The  zvagan  will  press  the  ashes  with  his 
hand.  Then  they  take  the  spathe  out  of  the  mosquito  bag 
(and  look  to  see  which  heap  has  been  disturbed)  and  they 
say,  "Yes,  it  is  his  father's  fault." 

This  extract  would  seem  to  indicate  that  ngglambi  infection 
comes  typically  down  the  patrilineal  line  and  illustrates  the 
identification  of  father  and  son. 

Case  2.  In  a  conversation  with  Tshava  of  Kankanamun 
I  enquired  whether  lando  (sister's  husband)  could  be  affected 
by  tawofitus  (wife's  brother's)  ngglambi.  My  informant  told 
me  that  such  infection  would  always  pass  to  the  lafido^  if  he 
gave  food  to  his  wife's  brother,  and  in  general  that  befriending 
a  sorcerer  was  dangerous  because  the  sorcerer's  ngglambi  was 
liable  to  affect  those  who  gave  him  food  or  lodging.  He  said 
that  if  any  man  saw  another  eating  a  particular  creeper  used 
in  sorcery,  he  would  spread  a  warning:  "Do  not  give  food 
to  that  man."  But  the  danger  is  only  attached  to  gifts  of  food 
and  not  to  gifts  of  valuables. 

^  It  appeared  from  Tshava 's  remarks  that  though  ngglambi  from 
tawontu  might  infect  lando,  the  reverse  was  not  the  case.  But  this  em- 
phasis upon  the  one  direction  may  have  been  due  to  the  experiences  of 
his  family  in  its  relationship  with  Malikindjin. 


Sorcery  and  Vengeance  59 

"  If  you  give  food  to  a  sorcerer  your  pig  may  die,  or  your 
dog  may  die,  or  your  child  may  die." 

In  illustration  of  this,  my  informant  quoted  the  case  of 
his  parents:  his  father  Djuai  had  married  Malikindjin's^ 
own  sister,  Nyakala.  She  was  always  giving  food  to  her 
brother  and  Djuai  often  scolded  her  for  this.  Because  she 
befriended  the  sorcerer  "many"  of  her  children  died,  and 
finally  she  herself  died.  Later,  Nauyambun,  a  very  old  woman, 
the  mother  of  Malikindjin,  came  to  the  house  and  my  in- 
formant was  constantly  sick  (because  of  this  connection  with 
the  sorcerer). 

Returning  to  the  question  of  his  mother's  death,  Tshava 
narrated  to  me  in  some  detail  how  he  and  Malikindjin  had 
avenged  her:^ 

When  Malikindjin  was  mourning  for  Nyakala  he  came  to 
me  and  said  to  me,  "We  will  go  down  to  such  and  such  a 
piece  of  bush."  We  went  and  there  found  a  man  named 
Tamwia.  Malikindjin  said,  "Now  we  are  going  to  take 
vengeance  for  your  mother  on  this  man."  He  told  me  to 
hide  myself  and  he  used  magic  to  make  Tamwia  unable  to 
see  him.  Tamwia  was  "in  the  dark"  and  Malikindjin  made 
him  "cold".  He  did  not  cry  out.  Malikindjin  came  straight 
up  to  him  and  broke  his  neck  and  head  with  an  adze.  He 
killed  him.  He  cut  off  his  head  and  put  in  its  place  a  head 
made  of  nggelakavwi  (a  tuberous  plant,  Myrmecodia  sp.). 
He  threw  the  real  head  away. 

We  each  took  an  arm  and  hid  the  body  in  the  elephant 
grass.    Malikindjin  said,  "Don't  talk  about  this,  but  there 

^  For  a  description  of  the  personality  of  the  old  sorcerer,  Malikindjin, 
cf.  p.  164. 

*  The  magical  incident  which  is  here  described  is  very  closely  com- 
parable with  one  collected  by  Dr  Fortune  in  Dobu.  In  both  incidents 
the  informant  acted  as  watch-dog  and  the  magic  was  of  the  type  called 
"vada".  In  it  the  victim  is  first  killed  and  then  brought  back  to  life  only 
to  die  later.  The  two  stories  differ  in  that  in  Dobu  it  is  the  heart  of  the 
victim  which  is  removed  instead  of  the  head,  but  I  was  told  on  another 
occasion  that  the  latmul  sorcerers  sometimes  remove  the  heart.  And  they 
differ  too  in  this,  that  Dr  Fortune's  informant  was  moved  almost  to 
hysteria  in  the  narration  of  the  story,  whereas  my  informant  described 
the  events  in  as  cool  and  as  detached  a  manner  as  he  would  have  described 
a  magical  procedure  for  the  improvement  of  yams.  (Cf.  Sorcerers  of 
Dobu,  1922,  pp.  158-164  and  284-287.) 


6o  Sorcery  and  Vengeance 

is  no  fear  of  trouble.  We  are  only  taking  vengeance."  Then 
I  went  away  to  work  on  Malikindjin's  garden,  and  came  back 
to  find  the  man  Tamwia  lying  sick  and  groaning.  Malikindjin 
woke  him  up  and  said,  "  Are  you  asleep? "  and  Tamwia  said, 
"Yes."  And  Malikindjin  said,  "  Can  you  see  me? "  Tamwia 
said,  "Yes."  Malikindjin  said,  "What  have  I  been  doing 
to  you?"  and  Tamwia  said,  "I  was  asleep  and  you  woke 
me  up."  Malikindjin  said,  "Right.  That  is  the  way  to  talk. 
Now  go  away." 

Malikindjin  also  told  him,  "You  cannot  cry  out  with 
sickness  soon.  You  will  stay  awhile  first",  and  he  gave  him 
a  date — five  days  thence. ^  "You  cannot  mention  me.  It  is 
not  good  for  you  to  live  in  Kankanamun.  You  had  better 
die."   Malikindjin  addressed  Tamwia  by  his  -awan  name.^ 

Then  we  went  back  to  the  village  and  Malikindjin  said  to 
Djuai,  "Do  not  wear  mourning  all  the  time,  I  have  taken 
vengeance.  You  can  take  off  your  mourning,  or  only  wear 
a  little."   The  poor  man,  Tamwia,  soon  died. 

After  this  Malikindjin  wanted  to  teach  me  his  spells  but 
I  was  afraid.  If  I  paid  him  with  valuables  and  they  were  not 
enough,  he  might  kill  me.  Later,  Malikindjin  debated  the 
matter  in  the  ceremonial  house.  He  said,  "My  sister  is 
balanced  by  a  man.  Why  did  you  send  your  wagan  to  kill 
my  sister?   But  now  we  are  avenged." 

But  Kwongu  sent  his  tshumhuk^  (personified  pointing  stick) 
to  kill  one  of  Malikindjin's  wives  and  she  died,  and  Malikindjin 
killed  Kwongu's  wife  and  said,  "You  killed  my  wife,  now 
I  avenge  it.  There  is  no  reason  why  I  should  hide  the  matter." 

Then  Kwongu  killed  a  child  of  Malikindjin  and  Malikindjin 
killed  the  wife  of  laremei.  laremei  complained  of  this  and 
said,  "Why  do  you  kill  my  wife?"  but  Malikindjin  said, 
"  Kwongu  killed  my  child  and  you  have  given  Kwongu  food. 
Why  do  you  give  him  food?    Now  his  trouble  has  come  to 

^  Among  the  latmul,  in  magical  transactions  when  a  date  is  fixed  it  is 
invariably  five  days  hence. 

^  I  had  to  ask  a  leading  question  to  get  this  detail.  But  I  was  on 
other  occasions  told  spontaneously  that  the  -awan  name  is  used  in  ad- 
dressing the  victim  of  sorcery,  or  his  image  (cf.  footnote,  p.  43). 

^  A  tshumbuk  is  a  pointing  stick,  generally  a  shortened  spear  shaft. 
It  is  personified  and  the  only  specimen  which  I  saw  had  a  face  worked 
in  clay  in  the  middle  of  the  shaft. 


Sorcery  and  Vengeance  6i 

you  and  your  wife  has  died."  laremei  said,  "True,  I  was 
stupid."  He  did  not  avenge  his  wife's  death  but  went  away 
to  Timbunke  village,  where  later  he  was  killed  in  war  by 
natives  of  Kararau. 

From  this  story  we  may  extract  the  following  identifica- 
tions, which  we  may  phrase  in  terms  of  ngglamhi: 

{a)  The  passage  of  ngglambi  from  the  sorcerer  to  his  sister's 
husband ;  i.e.  identification  of  a  man  with  his  sister  and  of  the 
sister  with  her  husband. 

(6)  The  passage  of  ngglambi  from  the  sorcerer  to  his 
mother  and  thence  to  those  who  befriended  her. 

{c)  The  passage  of  ngglambi  from  husband  to  wife  {A  kills 
5's  wife  and  B  kills  ^'s  wife). 

{d)  The  passage  of  ngglambi  from  father  to  son. 

{e)  The  passage  of  ngglambi  from  the  man  who  befriends 
the  sorcerer  to  his  wife. 

Case  3.  My  informant  Tshava  of  Kankanamun  gave  me 
another  instance  of  the  dangers  of  befriending  the  sorcerer. 
He  said  that  Tshe  always  gives  food  to  Malikindjin  and  that 
is  why  his  child  is  always  sick  and  many  pigs  have  been 
killed. 

Once  Malikindjin  said,  "  I  have  had  a  dream.  Somebody 
will  have  to  kill  a  pig"  (i.e.  he  had  dreamt  of  ngglambi  or 
some  disaster  and  a  pig  would  have  to  be  killed  to  propitiate 
the  wagan).  Then  there  was  a  debate  against  Malikindjin 
and  a  number  of  men  said,  "This  is  to  be  Tshe's  pig,  he 
is  going  to  pay  now."  And  they  said  to  Tshe,  "You  are  always 
giving  food  to  Malikindjin  and  now  you  will  get  his  trouble." 
And  Djuai  (Malikindjin's  sister's  husband)  said,  "Always 
Malikindjin's  sister  caused  trouble  in  my  house,  and  now 
Malikindjin  has  gone  to  your  house  and  your  children  will 
die.  If  anyone  kills  a  pig  you  will  have  to  pay  for  it"  (i.e. 
there  was  a  danger  that  if  to  avoid  the  disaster  somebody  had 
to  kill  a  pig,  Tshe  would  be  held  partly  to  blame ;  and  there- 
fore unless  he  paid  for  the  pig  his  children  would  die).  But 
Tshe  said,  "  Malikindjin  is  my  father  (classificatory) ;  I  cannot 
not  help  him."i 

^  The  avoidance  between  father  and  son,  whereby  the  father  must 
never  eat  food  produced  by  the  son,  is  only  observed  by  own  father  and 


62  Sorcery  and  Vengeance 

In  this  instance,  we  have  a  very  clear  illustration  of  the 
passage  of  ngglamhi  to  a  man  who  befriends  the  sorcerer 
because  he  regards  him  as  a  "father".  The  man  is  even 
expected  to  pay  in  advance  for  the  effects  of  the  sorcerer's 
evil  deeds  in  order  to  ward  off  the  danger  which  will  other- 
wise descend  upon  his  own  children. 

Case  4.  This  is  a  case  of  identification  of  husband  and  wife 
of  which  I  heard  when  collecting  census  material  in  Palimbai. 

The  husband  had  assisted  the  Australian  government  in 
putting  down  head-hunting.  He  and  another  man  had  used 
their  influence  to  make  the  natives  of  Palimbai  burn  their 
spears.  Later,  his  wife  died  and  her  belly  was  much  swollen. 
They  found  a  piece  of  a  burnt  spear  in  it.  (Probably  this 
fragment  was  produced  in  a  conjuring  trick  performed  by 
a  medical  practitioner.)  This  was  regarded  as  proof  that  she 
had  died  from  sorcery  resulting  from  the  burning  of  the 
spears. 

Case  5.  Informant:  Gimelwan  of  Palimbai.  A  quarrel 
arose  between  Djuai^  and  Pandu  over  the  tidying  of  the 
ceremonial  house.  Djuai  was  sweeping  the  ceremonial  house 
and  picking  up  the  rubbish  with  a  pair  of  planks.  He  was 
cross  with  Pandu  because  he  was  doing  nothing.  Pandu 
kicked  Djuai  and  Djuai  then  killed  Pandu  "  with  a  small  stick 
and  broke  his  belly".   Pandu  was  buried. 

Lapndava,  the  son  of  Pandu,  threw  a  spear  into  Djuai's 
house  (a  gesture  of  anger  and  challenge).  Djuai  then  pre- 
sented twenty  shell  valuables  of  various  sorts  to  Lapndava. 

But  the  matter  did  not  finish  there.  Later,  Djuai's  "  children 
died  and  his  wife  and  now  he  is  all  alone"  (the  implication 
being  that  the  wife  and  children  died  from  sorcery  following 
upon  the  quarrel). 

On  the  other  side  Lenagwan,  Pandu's  elder  brother,  was 
pursued  by  misfortune.  "He  curled  his  lip  in  Kankanamun" 
(i.e.  he  sulked).  He  wanted  to  marry  a  woman  in  Kan- 
father's  tambinyen.  It  is  correct  for  a  man  to  give  food  to  a  distant 
classificatory  father  (in  this  case,  father's  mother's  sister's  son).  More- 
over the  taboo  is  relaxed  when  the  father  is  very  old. 

^  The  Djuai  of  this  story  was  a  man  of  Palimbai,  not  the  father  of 
Tshava  who  has  the  same  name  and  is  mentioned  in  the  other  stories. 


Sorcery  and  Vengeance  63 

kanamun  but  could  not.  Then  he  went  to  the  debating  stool 
in  the  Kankanamun  ceremonial  house,  and  called  the  men 
out  to  raid  Kararau,  and  they  went  out  to  shoot  the  women 
while  they  were  fishing  on  the  Kararau  lake.  Lenagwan  threw 
his  spears  but  killed  no  one  and  the  raid  was  a  failure.  Later, 
the  people  of  Kararau  organised  another  raid  in  reply  and 
killed  three  women  of  Kankanamun,  and  finally  Lenagwan 
was  killed  by  sorcery  done  by  the  men  of  Kanakanamun 
because  he  had  organised  the  ineffectual  raid.  "Therefore 
Pandu's  clan  is  now  almost  finished."  And  my  informant 
mentioned  two  other  members  of  the  clan  who  had  died,  one 
of  them  killed  by  Europeans  and  the  other  killed  by  sorcery. 
This  case  contains  identifications  of  father  and  son,  man 
and  wife,  elder  brother  and  younger  brother,  and  a  vague 
reference  to  identification  based  upon  clan  membership  at 
the  end  of  the  story. 

Case  6.  Informant:  Waindjamali  of  Palimbai.  This  story 
was  given  me  in  explanation  of  a  shamanic  seance  which  I  had 
recently  witnessed.  The  seance  was  held  to  enquire  into  the 
death  of  Tepmanagwan,  the  great  man  whose  mortuary  image 
is  described  later  (p.  154).  The  death  was  the  latest  result 
of  a  feud  which  had  started  between  rival  initiatory  grades 
some  four  or  five  years  earlier.  My  informant  was  a  member 
of  grade  By  4  (cf.  Diagram,  p.  245);  Tepmanagwan  was  a 
member  of  By  2  and  the  quarrel,  which  started  between 
Ax  3  and  By  4,  spread  to  include  almost  the  whole  of  Ax 
and  By. 

The  bandi  {Ax  3)  called  us  (kamberatly  By  4)  out  to  fight. 
They  beat  on  the  slit-gongs  and  insulted  our  fathers.  "You 
children  are  trash  and  bastards.  We  have  played  with  your 
mothers."  Then  there  followed  a  brawl  with  no  spears  but 
only  sticks.  We  were  beaten,  but  we  said,  "We  will  avenge 
it  on  their  children"  {Ax  5).  Then  the  other  moiety  made 
us  buy  pigs  and  we  presented  them.  Kumbwi  (a  man  of 
Ax  i)  and  his  brother  Malishui  were  conspicuous  in  this 
brawl.   Kumbwi  beat  a  slit-gong^  to  call  his  moiety  together 

^  In  descriptions  of  brawls,  it  is  usually  stated  who  beat  the  gongs; 
and  from  Case  7  it  would  appear  that  the  man  who  beats  the  gongs  is 
considered  responsible  for  the  brawl. 


64  Sorcery  and  Vengeance 

and  carried  a  spear.  Malishui  only  carried  a  stick.  Then 
Tepmanagwan  came  up  and  contradicted  Kumbwi.  Tep- 
managwan  picked  up  a  spear  with  barbs  (such  as  nominally 
should  not  be  used  in  brawls  inside  the  village)  and  he 
speared  Malishui  in  the  leg.  Then  all  the  opposite  moiety 
combined  against  Tepmanagwan,  but  his  own  moiety  came 
to  his  help  and  he  was  also  helped  by  his  classificatory  laua 
and  his  own  daughter's  husband.  After  the  brawl  Tep- 
managwan said,  "So  far  they  have  always  bullied  us;  to- 
morrow we  shall  fight.  So  far  we  have  always  used  bamboo 
points;  to-morrow  we  shall  use  barbs."  Next  day  there  was 
a  brawl  and  Tepmanagwan  was  wounded  "  with  many  spears  " 
and  that  night  we  all  scattered.  I  and  others  of  the  Kam- 
berail  grade  {By  4)  went  to  Tegowi  village.  Tepmanagwan 
went  to  Malikindjin,  his  classificatory  brother  in  Kankanamun. 
He  gave  a  mother-of-pearl  crescent  and  conus  disks  to  the 
old  sorcerer — and  Malishui  died. 

Soon  after  that  Wandem  (Tepmanagwan 's  younger  brother's 
wife)  died.  "Malikindjin  killed  her  with  a  tshumbuk  (per- 
sonified pointing  stick)  as  retribution  for  Malishui.  He  killed 
her  'in  return'  because  he  was  afraid  they  would  turn  the 
ngglamhi  back  on  to  himself  or  his  relatives." 

Then  Kumbwi  wept  and  gave  valuables  to  his  wagan 
(ancestral  spirit)  and  said,  "Ancestor,  my  younger  brother 
only  has  died,  but  they  shall  not  stay  unhurt." 

Then  two  sons  of  Tepmanagwan  died,  one  of  them  while 
he  was  away  as  an  indentured  labourer. 

Then  there  followed  five  other  deaths. ^  I  tried  to  get 
information  as  to  the  exact  relevance  of  these  deaths  to  the 
feud,  but  my  informant  dismissed  the  matter,  "They  are  not 
straight.  They  only  turn  about.  That  is  Malikindjin's  way." 

Finally  Kumbwi  gave  a  mother-of-pearl  present  to  his 
loagan  and  Tepmanagwan  died ;  and  in  the  seance  enquiring 
into   his   death  the  zcagan  (i.e.   the  shaman  in  a  state  of 

^  Probably  most  of  the  deaths  in  this  sorcery  feud  were  due  to  in- 
fluenza. But  we  are  here  concerned,  not  with  the  theories  of  European 
medicine,  but  with  those  of  the  latmul.  In  general  no  deaths  are  regarded 
by  the  latter  as  "natural"  but  all  are  referred  to  sorcery.  Even  when  a 
man  is  killed  by  the  enemy,  his  relatives  will  suspect  that  some  sorcerer 
in  the  village  has  sold  the  dead  man's  soul  to  the  enemy  before  the  raid. 


Sorcery  and  Vengeance  65 

possession)  produced  this  mother-of-pearl  crescent  and  said, 
"  This  is  the  shell  that  killed  Tepmanagwan  ",  and  he  smashed 
it  against  the  post  of  the  ceremonial  house  as  a  sign  that  that 
feud  was  finished. 

This  case  contains  first  of  all  an  example  of  solidarity 
within  the  initiatory  moieties,  but  the  ties  between  laua  and 
zoau  and  between  son-in-law  and  father-in-law  are  given 
preference  over  allegiance  to  the  moiety.  In  the  sorcery  feud 
we  have  identifications  of  brother  and  brother,  of  younger 
brother's  wife  and  husband's  elder  brother,  and  of  sons  and 
father. 

Case  7.  This  is  the  story  of  a  long  sorcery  feud  in  Kan- 
kanamun  which  was  given  me  by  my  informant,  Tshava. 
As  in  the  case  of  the  similar  feud  in  Palimbai,  this  series  of 
deaths  had  its  beginning  in  a  quarrel  between  initiatory 
moieties : 

A  boy  was  playing  the  Pan-pipes  in  the  junior  ceremonial 
house  and  a  woman,  lalegwesh,  came  up  outside  the  house. 
The  boy  said,  "  Come  up.  I  want  to  see  your  sore."^  She 
went  away  but  later  came  back  to  the  junior  ceremonial  house 
and  shouted,  "  My  windjimbu,^  I  want  to  see  you."  Then  the 
boy  hid  and  went  to  his  house.  Later  lalegwesh  was  heard 
playing  the  Pan-pipe  tunes  on  a  jews'  harp,  and  the  boy  came 
with  others  and  scolded  her.  She  said,  "What,  do  you  think 
I  don't  know  the  Pan-pipe  tunes?  Mwaimali  (her  husband) 
taught  me  them."  A  senior  man,  Katka,  overheard  this  and 
was  angry.  He  went  off  and  told  the  other  men  and  chased 
lalegwesh  with  a  stick.  He  found  some  men  beating  on  a 
suspended  log.^  He  said,  "  Stop  doing  that.  A  woman  has 
exposed  all  our  secrets."  Then  Djindjime  beat  a  summons 
on  the  slit-gong  and  all  the  men  came  together.  They  went 
and  raided  lalegwesh's  house  and  broke  her  pots;  but  her 
brothers  came  to  her  assistance  and  there  was  a  brawl  between 
them  and  Katka.   At  first  they  were  many  and  he  was  only 

^  An  obscene  metaphor  for  female  genitals. 

^  The  various  secret  sound-producing  objects,  flutes,  Pan-pipes,^  etc. 
are  regarded  as  connected  with  the  wood  spirits  or  windjimbu.  There  are 
many  stories  of  women  who  have  had  love  affairs  with  windjimbu. 

^  This  is  a  junior  analogue  of  the  wagan  slit-gongs,  an  initiatory  secret. 
The  tunes  for  the  Pan-pipes  are  also  secret. 


66  Sorcery  and  Vengeance 

helped  by  the  small  boys  (of  the  junior  ceremonial  house). 
Katka  fell  in  the  brawl  and  then  others  came  to  his  assistance. 
His  eye  was  broken. 

Next  day  Malikindjin  came  and  saw  Katka's  eye  and  asked 
what  had  happened,  Katka  said,  "We  had  a  brawl."  Mali- 
kindjin said,  "  Why  ? "  and  Katka  said,  "  About  men's  things  " 
(i.e.  initiatory  secrets).  Then  they  debated  the  matter  and 
Malikindjin  took  the  side  of  lalegwesh  and  Mwaimali.  He 
said,  "  Somebody  will  die  from  this  trouble."  Later,  he  sent 
his  tshumbuk  (personified  pointing  stick)  and  killed  Kwongkun. 

On  the  day  after  this  death,  some  young  men  came  home 
from  indentured  labour  with  their  boxes  full  of  trade  goods. 
Malikindjin  went  to  the  ceremonial  house  and  demanded  a 
box  of  trade  goods  from  the  labourers.  "  I  have  been  looking 
after  your  women  while  you  were  away.  You  must  pay  me 
one  box."  The  boys  paid  him,  but  they  said,  "That  box 
will  eat  up  a  man  later," 

The  box  ate  Katka  (i,e.  Katka  died  of  sorcery  because  it 
was  against  him  that  lalegwesh  had  needed  Malikindjin's 
help).  Katka's  death  was  avenged  on  loli  and  his  younger 
brother,  Kwilegi,  who  had  helped  lalegwesh.  These  two  died 
in  one  night  (this  was  probably  during  an  influenza  epidemic). 
Later,  Katka's  wife  died  to  balance  Kwilegi.  Then  the  wagan 
said,  "Who  else  helped  Katka?"  and  Djindjimowe  died,  and 
to  balance  him  Kisak. 

Then  the  wagan  said,  "Who  beat  the  slit-gongs?"  and 
Djindjime  died.  Finally  lalegwesh  died,  killed  by  Mali- 
kindjin's spells.  The  old  sorcerer  was  acting  on  both  sides. 
The  beginning  of  the  trouble  lay  with  her  and  that  is  why 
she  died,  and  they  all  said,  "That  talk  is  finished.  It  is 
done  with," 

But  later  Mwaimali,  lalegwesh's  husband,  married  Kagan, 
the  widow  of  Katka,  and  she  bore  him  a  child.  Then  Mali- 
kindjin's wagan,  Tshugu,  said,  "Yes,  at  first  a  woman  made 
trouble  and  Katka  died.  Now  you  have  married  Katka's 
wife."   Then  the  child  died. 

Then  Mwaimbwan,  a  member  of  Mwaimali's  clan,  said, 
"This  affair  was  done  with  long  ago,  but  now  you  want  to 
start  it  again  "  and  he  caused  a  child  to  die  on  the  other  side. 


Sorcery  and  Vengeance  67 

Finally  there  was  a  seance.  Malikindjin  went  into  a  trance 
and  Tshugu  (speaking  through  Malikindjin's  mouth)  called 
upon  all  who  had  been  concerned  in  this  feud  to  kill  a  big 
pig  to  finish  the  feud.  He  asked  for  the  names  of  all  who  had 
taken  part  in  the  original  brawl  on  each  side,  and  told  the 
members  of  both  sides  to  contribute  valuables.  A  pig  was 
bought  and  killed  and  the  various  tvagan  (in  their  shamans) 
and  all  the  men  ate  it.  After  the  feast  all  the  wagans  took  the 
bones  of  the  pig  and  they  made  a  hole  close  to  the  ladder  of 
the  ceremonial  house  and  there  buried  them,  at  the  sug- 
gestion of  Tshugu;  and  they  said,  "If  any  man  raises  this 
feud  again,  he  shall  die." 

But  later  Malikindjin  fell  sick.^  That  is  because  everybody 
wants  vengeance  upon  him.  All  who  know  anything  about 
spells  and  all  the  wagan,  they  are  all  agreed. 

In  this  story  the  following  identifications  are  recognisable : 

{a)  Identifications  based  either  on  sex  solidarity  or  on 
membership  of  initiatory  moieties  (cf.  the  case  quoted  on 
p.  99,  in  which  a  woman  was  detected  spying  upon  the 
flutes). 

{h)  A  woman's  brothers  side  with  her,  giving  precedence 
to  that  allegiance  over  their  identification  with  their  sex  or 
moiety. 

{c)  Those  who  help  the  offending  woman  suffer,  including 
a  pair  of  brothers. 

{d)  Katka's  wife  dies  from  her  husband's  ngglamhi. 

{e)  Finally  the  marriage  of  Mwaimali  to  Katka's  widow 
provides  such  a  tangled  position  that  it  is  not  clear  whether 
the  child  that  was  born  to  this  marriage  died  because  of 
lalegwesh's  sins  or  those  of  Katka. 

Case  8.  Tshava  of  Kankanamun  told  me  the  following 
story  about  the  death  of  two  of  his  father's  wives,  Mait  and 
Undemai : 

Two  of  Malikindjin's  wives,  Mwaim  and  Lemboin,  were 
quarrelling.  Mwaim  said,  "  Djuai  (my  informant's  father)  is 

^  He  died  about  four  months  later.  I  enquired  about  the  burying  of 
the  bones  of  the  pig  and  was  told  that  this  was  not  a  customary  pro- 
cedure, but  was  done  simply  at  the  wagon's  suggestion. 


68  Sorcery  and  Vengeance 

always  playing^  with  you."  Malikindjin  heard  of  this  and 
fetched  help  from  Tegowi  village  to  avenge  himself  on 
Djuai ;  and  there  was  a  brawl  in  Kankanamun.  Later,  Djuai 
killed  a  big  pig  and  gave  it  to  the  men  of  Palimbai,  inviting 
them  to  come  and  kill  lowimet,  a  native  of  Tegowi  who  was 
settled  in  Kankanamun  (lowimet  and  Malikindjin  were  classi- 
ficatory  brothers).  But  lowimet  fled  to  Tegowi  and  they 
only  destroyed  his  property  and  stole  his  shell  valuables. 
Later  he  came  back  and  wept  for  his  things.  He  had  a  string 
bag,  a  melo  shell,  a  Conus  shell  necklace  and  a  Turbo  shell. 
These  he  gave  to  Malikindjin.  Malikindjin  said,  "Why  do 
you  give  them  to  me? "  lowimet  said,  "  I  am  crying  for  my 
things.  They  have  killed  my  pig,  etc.  I  want  one  of  their 
women  to  die."  Then  the  two  of  them  agreed  on  a  day — "  on 
this  day  we  will  meet  in  the  bush". 

On  the  date,  they  met  and  they  waited  for  Mait  (Djuai's 
wife)  at  the  place  called  Tungwimali  on  the  river  bank.  When 
she  came  they  lied  to  her  and  said,  "There  are  crocodiles' 
eggs  here ",  and  when  she  came  ashore  they  killed  her — 
killed  her  completely  and  revived  her.  She  went  back  and 
told  my  father  (Djuai),  "It  was  those  two  who  killed  me. 
They  lied  about  crocodiles'  eggs  and  I  went  ashore."  Djuai 
only  heard  what  she  said  and  then  she  died.  Djuai  wept  for 
her  and  called  upon  his  wagan,  Tshugu.^  He  said,  "  I  am  the 
only  one  who  is  weeping.  They  are  untouched.  Later  they 
must  cry  too."  Then  Tshugu  killed  Mwaim,  the  wife  of 
Malikindjin. 

Later  there  was  a  seance  and  Tshugu  possessed  Malikindjin 
and  spoke  addressing  Malikindjin,  "You,  my  shaman,  why 
are  you  mad  and  kill  all  the  women?  Now  they  are  taking 
vengeance.  My  clansman  called  on  me  and  I  came  and  killed 
the  woman.  One  of  your  wives  has  died.  Later  somebody 
belonging  to  lowimet  shall  die."  Then  Yivet,  lowimet's 
daughter's  husband,  died.  Then  Malikindjin  and  lowimet 
started  again  and  killed  Undemai,  another  wife  of  my  father. 
Then  it  was  all  balanced. 

^  A  euphemism  for  copulation. 

^  This  was  the  wagan  which  possessed  Malikindjin.  My  informant 
explained  that  when  Djuai  called  upon  Tshugu  it  was  only  the  spirit 
who  heard,  not  the  shaman. 


Sorcery  and  Vengeance  69 

But  lowimet  did  spells  on  an  adze  and  found  one  of  my 
footprints.  He  hit  the  ground  with  the  adze  and  then  wrapped 
up  the  adze  in  nettles  and  put  it  in  a  pot  and  made  it  hot  on 
the  fire.  My  sore  hurt  me  and  I  cried  out.  Then  lowimet 
put  a  little  water  on  the  adze  and  cooled  it,  and  my  sore  was 
cooled.  This  went  on  till  I  took  a  Turbo  shell  and  my  wife 
carried  it  to  lowimet.  Then  lowimet  said,  "You  have  paid 
me  and  later  the  sore  will  dry  up."  It  dried  a  little,  but  the 
doctor  (i.e.  the  Government  medical  official)  took  me  up  to 
Ambunti  (to  the  Government  hospital).  The  doctor  said, 
"Why  is  the  sore  so  big?"  And  I  said,  "  It  is  only  lowimet." 
The  doctor  said,  "What  was  the  quarrel  about? "  And  I  said, 
"  It  was  my  father's  quarrel,  but  it  is  not  right  for  the  son 
to  help  his  father."^  Then  the  doctor  sent  a  police  boy  to 
get  lowimet  and  asked  him,  "Why  did  you  hurt  this  boy?" 
lowimet  said  he  did  not  know,  and  the  doctor  said,  "Tshava 
has  told  me.  Why  are  you  lying? " 

lowimet  was  nearly  put  in  prison,  but  he  is  my  (classi- 
ficatory)  wau.  The  doctor  asked  me  if  I  was  sorry  for  him 
and  I  said  "Yes",  and  the  doctor  said  "Never  mind"  and 
dismissed  the  matter. 

In  this  story  the  following  identifications  occur: 

{a)  Malikindjin  and  lowimet;  brother  with  classificatory 
brother. 

{b)  Malikindjin  and  Mwaim;  husband  and  wife. 

{c)   Djuai  and  Mait ;  husband  and  wife. 

{d)  lowimet  and  his  daughter's  husband. 

{e)   Djuai  and  Undemai ;  husband  and  wife. 

(/)  Djuai  and  Tshava;  father  and  son. 

{g)  And  finally  Tshava  "  is  sorry  for  "  his  classificatory  wau. 

Case  9.  This  story  was  given  me  by  Tshava  of  Kankana- 
mun.  It  describes  the  vengeance  of  the  wagan  upon  a  medical 
practitioner  who  interfered  in  their  affairs  and  demanded  no 
payment  for  his  services : 

Mongwaevi,  the  wagan  of  Yivom's  clan,  had  told  him  to 
kill  a  fowl,  but  Yivom  said,  "Why  should  I?   When  should 

^  Cf.  the  taboo  upon  the  father's  eating  any  food  grown  or  prepared 
by  the  son. 


yo  Sorcery  and  Vengeance 

I? "  and  refused.  Then  the  zvagan  turned  and  made  Yivom's 
small  son  sick,  and  Tengai,  a  man  with  some  knowledge  of 
spells,  cured  him.  But  the  wagan  said,  "It  was  not  his 
(Tengai's)  affair.  It  was  a  case  of  which  people  had  said  *  he 
is  going  to  die',  but  that  man  Tengai  used  spells." 

Then  the  wagan  made  Tengai's  wife  sick.  The  people  said, 
"Yes,  they  paid  Tengai  nothing  for  his'spells  and  he  did  not 
ask  for  pay  (i.e.  inasmuch  as  he  had  not  been  paid  it  was  not 
his  business  to  interfere)."  Punpunbi's  daughter  said,  "It 
is  witches"  and  "Yivom's  folk  will  not  stay  untouched.  Our 
zvagan  will  look  at  them.  It  is  their  ngglambi."  Punpunbi 
agreed  with  her  daughter's  opinion  and  she  summoned  her 
husband  Tengai  and  was  cross  with  him.  She  sent  him  away 
from  her;  though  before  that  they  had  agreed  well  enough 
together.  For  a  long  while  they  did  not  summon  the  wagan, 
but  finally  they  called  for  Mongwaevi's  shaman  who  went 
into  a  state  of  possession  and  the  zvagan  demanded  areca 
nuts  from  Tengai.  Only  then,  after  the  presentation  of 
areca  nuts,  did  Mongwaevi  clear  the  road  to  the  land  of  the 
dead.  The  soul  had  wanted  to  go  before  but  Mongwaevi 
(in  spite)  was  closing  the  road.  The  woman  was  terribly  sick 
but  could  not  die.  Then  she  died. 

In  this  case  Yivom's  original  offence  against  Mongwaevi 
is  first  avenged  upon  his  child  (identification  of  father  and 
son) ;  then  the  ngglambi  passes  to  Tengai — an  example  of  the 
danger  of  helping  those  with  ngglambi.  Then  from  Tengai 
the  ngglambi  passes  to  his  wife. 

Case  ID.  This  story  was  related  to  me  by  Tshava  of 
Kankanamun.  I  have  rearranged  the  story  so  that  the 
sequence  of  events  may  be  clear : 

Atndjin  married  a  woman  named  Tshugwa,  and  later  she 
helped  him  to  marry  her  younger  sister,  Wama,  by  con- 
tributing to  the  bride  price  valuables  which  she  had  got  by 
breeding  pigs.  My  informant  told  me  that  it  is  quite  usual 
for  an  elder  sister  to  help  her  husband  to  marry  her  younger 
sister. 

Later  Atndjin  died,  leaving  the  two  sisters  as  widows. 
According  to  custom  they  should  have  become  wives  of 


Sorcery  and  Vengeance  yi 

Atndjin's  younger  brother,  but  this  did  not  occur.  A  man 
named  Tshuat  of  Jentschan  wanted  to  marry  the  elder  sister 
and  finally  took  her  to  Jentschan,  but  Malikindjin  wanted 
her  and  sent  his  tshumbuk  (personified  pointing  stick).  Tshuat 
saw  the  tshumbuk  in  a  dream  and  was  frightened,  so  he  let 
her  go.  Malikindjin  took  her  and  married  her  and  paid  a 
bride  price  to  Atndjin's  younger  brother.  Then  Malikindjin 
said,  "When  the  younger  sister  marries  I  shall  get  part  of 
the  bride  price — as  much  as  the  elder  sister  contributed — 
never  mind  about  a  pig — or  if  there  are  not  enough  valuables, 
let  them  pay  a  pig." 

But  the  various  husbands  of  the  younger  sister  endeavoured 
to  dodge  this  payment.  First  a  man  named  Tevwa  took  her, 
but  he  became  sick  and  was  frightened  of  Malikindjin  and 
sent  her  away.  For  a  while  she  remained  unattached  till 
Tshaunanti  took  her.  But  his  other  wife  died  because  of  her ; 
Malikindjin  sent  his  tshumbuk  because  the  pay  was  inadequate. 
They  said,  "Tshugwa's  things  (i.e.  valuables)  have  eaten  up 
your  wife."  Then  Tshaunanti  sent  her  away.  Then  Mbulm- 
buangga  married  her,  but  his  child  by  a  previous  wife  died 
because  he  did  not  kill  a  pig  for  Malikindjin.  Though  his 
child  had  died  he  still  did  not  pay  the  sorcerer.  Malikindjin 
was  an  old  man  and  Mbulmbuangga  said,  "He  is  no  good 
any  more.  He  used  to  be  a  hot  man,  but  he  is  no  good  any 
more." 

In  this  case  we  see  first  a  husband,  Malikindjin,  identifying 
himself  with  his  wife  and  on  that  ground  demanding  recom- 
pense for  valuables  which  she  had  expended.  Later,  we  see 
the  offences  of  successive  husbands  of  the  younger  sister 
being  avenged  (i)  on  the  husband  himself,  (2)  on  the  hus- 
band's other  wife,  and  (3)  on  the  husband's  child  by  a 
previous  wife. 

Case  II.  This  is  a  story  which  was  told  me  by  Tshava 
of  Kankanamun.  The  story  is  irrelevant  to  the  question  of 
identifications  with  which  we  are  immediately  concerned,  but 
it  is  of  interest,  after  this  series  of  feuds,  to  cite  a  case  in 
which  a  man  killed  his  wife  and  afterwards  received  homicidal 
ornaments  from  her  relatives : 


ya  Sorcery  and  Vengeance 

Laindjin  killed  Tualesh,  his  wife.  She  was  a  member  of 
Wolgem  clan  and  (therefore)  my  (classificatory)  sister. 

Tualesh  went  to  get  water  in  a  long  bamboo.  When  she 
returned  Laindjin  said,  "Why  were  you  so  long? "  She  said, 
"I  was  not  long.  What  are  you  thinking?"  Laindjin  said, 
"I  was  only  asking,"  and  she  said,  "Other  men  don't  ask, 
but  you  do.  You  are  jealous."  Laindjin  said,  "Yes,  I  know 
how  women  carry  on."  Then  he  jumped  up  and  seized  the 
water  bamboo  and  bashed  her  with  it  and  she  died.  They 
made  incisions  in  her  skin  with  a  bamboo  knife  and  used 
spells,  but  they  could  not  cure  her.  Then  Laindjin 
wept. 

The  members  of  Wolgem  clan  heard  of  it  and  we  said, 
"Our  sister  is  dead."  Then  we  took  spears  and  went  to 
Laindjin's  house.  He  had  blocked  the  entrance  (probably 
with  a  mat)  and  was  inside.  Then  we  flung  many  spears  at 
the  house  and  Laindjin's  brother's  son  was  wounded,  but 
not  killed. 

We  presented  tamhointsha  (homicidal  tassels),  black  paint 
and  Crinum  leaves  to  Laindjin  and  he  killed  a  pig  for  the 
wagan  and  the  other  men,  but  we  did  not  eat  it;  it  was  our 
sister's  pig. 

Later,  Laindjin  said,  "This  woman  did  not  die  without 
a  cause.  Malikindjin  killed  her."  And  he  called  for  Mali- 
kindjin  and  said,  "Let  us  fight  with  spears."  Malikindjin 
fixed  a  day,  "the  day  after  to-morrow  I  shall  come  at  mid- 
day." Malikindjin  brought  a  big  bundle  of  spears  and  said, 
"Have  you  dreamt  of  me?  Did  I  kill  the  woman?"  and 
Laindjin  said,  "Yes.  I  sent  my  tshumbuk  (to  investigate). 
The  tshumbuk  woke  me  up  and  told  me — '  Malikindjin  killed 
her  for  no  reason.  The  woman  had  no  ngglambi.  She  was 
not  killed  by  many  men.  It  was  only  Malikindjin'."  Then 
Malikindjin  said,  "Come  on,  let  us  fight,"  and  Laindjin 
said,  "I  am  waiting  for  you.  Then  we  shall  fight  to- 
gether." 

The  two  fought  hard  and  their  lanoa  (sister's  husbands, 
etc.)  tried  to  hold  them  back  but  they  could  not.  Then 
Laindjin  was  wounded  in  the  hand  and  the  lanoa  managed 
to  hold  him. 


Sorcery  and  Vengeance  73 

In  compensation  1  for  the  death  of  the  woman  Laindjin 
paid  one  mother-of-pearl  crescent,  one  tortoise-shell  arm 
band,  two  Conus  shell  necklaces  and  three  Turbo  shells,  in 
a  string  bag  which  was  hung  up  in  the  Wolgem  ceremonial 
house. 

This  series  of  stories  contains  about  forty  identifications, 
but  such  a  series  is  by  no  means  sufficient  to  give  any  estimate 
of  the  relative  frequency  and  importance  of  the  various  types 
of  identification.  The  series  serves  only  as  a  set  of  illustra- 
tions of  the  working  of  identification  in  these  contexts.  It  is 
worth  remarking,  however,  that  a  rather  larger  series  of 
stories,  properly  documented  with  their  kinship  setting, 
would  give  enough  identifications  to  call  attention  to  the 
factors  which  influence  the  distribution  of  identification. 

^  It  is  not  clear  whether  the  presentation  of  homicidal  ornaments  pre- 
ceded this  compensation  or  followed  it.  In  other  cases,  it  appears  that 
the  homicide  pays  compensation  before  the  relatives  give  him  tambointsha. 


CHAPTER  VI 

Structural  Analysis  of  the  Wau-laua  Relationship 

WE  have  seen  that  the  term  "mother"  is  used  for 
mother's  brother  in  the  naven  ceremonies.  This  use 
is  not  confined  to  this  ceremony;  it  occurs  too  in 
initiation,  and  to  some  extent  in  ordinary  conversation.  Thus 
certain  male  natives  who  regarded  me  as  their  classificatory 
**  sister's  child  "  on  the  basis  of  a  matrilineal  name  which  had 
been  given  to  me,  used  occasionally  to  refer  to  themselves 
as  my  "mothers"  when  stressing  their  kinship  connection 
with  me.  One  of  them  would  even  say  to  me :  "You  are  the 
child  that  we  bore."  ("We  "  here  refers  to  the  maternal  clan.) 
At  other  times  informants  would  refer  to  the  clan,  whose 
matrilineal  name  I  bore,  with  such  a  phrase  as:  "They  are 
your  mother."^ 

Thus  it  is  clear  that  identification  of  the  mother's  brother 
with  the  mother  is  very  definitely  formulated  in  the  termino- 
logy, and  from  the  linguistic  usage  we  may  expect  to  find 
that  the  relationship  with  the  whole  maternal  clan  is  coloured 
by  the  relationship  to  the  mother  and  that  the  resemblance 
to  a  mother  is  most  strongly  marked  in  the  case  of  a  wau. 
But  we  need  not  expect  to  find  that  the  whole  of  a  mother's 
brother's  behaviour  can  be  described  as  the  behaviour  of  a 
male  mother.^  I  mentioned  above  that  a  man  is  to  some 
extent  identified  with  his  father,  and  we  may  expect  to  find 
that  this  identification  has  some  eff^ect  on  the  mother's  brother 
relationship.  The  laua  may  be  to  some  extent  identified  with 
his  father,  and  in  this  case  we  shall  find  the  wau  behaving 
towards  him  as  if  he  were  the  wau's  sister's  husband.  We  may 
therefore  expect  the  wau's  behaviour  to  fall  into  at  least  two 
categories  according  as  he  is  acting  upon  the  brother-sister 

^  The  language  has  no  plural  form  for  nouns  so  that  it  is  not  certain 
whether  the  word  "mother"  in  this  sentence  should  be  plural  or  singular. 

*  In  this  analysis  I  am  much  indebted  to  RadclifFe-Brown's  classic 
paper  on  the  Mother's  Brother  {South  African  Journal  of  Science,  1924, 
P-  542). 


Structural  Analysis  of  Wau-laua  Relationship     75 

identification  and  treating  the  laua  as  a  mother  would  treat 
her  child,  or  upon  the  father-son  identification  and  treating 
the  laua  as  a  man  would  treat  his  sister's  husband. 

The  same  considerations  apply,  mutatis  mutandis,  to  the 
behaviour  of  the  laua,  who  may  therefore  be  expected  to  treat 
his  wau  both  as  a  mother  and  as  a  wife's  brother.  Lastly 
it  is  possible  that  the  wau-laua  relationship  has  features  which 
are  due  to  other  identifications  which  we  have  not  considered, 
or  that  the  relationship  may  have  certain  features  sui  generis. 

Thus  we  are  led  to  an  experimental  analysis  of  the  wau- 
laua  relationship  under  the  three  following  heads : 

1 .  Behaviour  which  is  typical  of  the  relationship  between 
a  mother  and  her  child. 

2.  Behaviour  which  is  typical  of  the  relationship  between 
a  wife's  brother  and  a  sister's  husband. 

3.  Other  details  of  behaviour  which  do  not  fit  either  of 
these  patterns. 

By  working  with  these  three  categories  we  shall  be  able  to 
set  out  all  the  available  information  about  the  wau-laua  rela- 
tionship in  a  scheme  which  will  show  clearly  how  much  or 
little  of  a  wau's,  and  laua's,  behaviour  is  to  be  regarded  as 
sui  generis  and  how  much  is  based  upon  these  identifications. 

I.  The  latmul  mother's  behaviour  towards  her  offspring 
is  simple,  uncomplicated  by  self-contradictory  elements  such 
as  characterise  the  patterns  of  a  father's  behaviour.  Her  acts 
and  attitudes  may  be  seen  as  built  up  around  one  central 
cultural  act;  she  feeds  the  child,  first  suckling^  it  and  later 
giving  it  food.  It  is  evident  too  from  the  behaviour  of  the 
wau  that  the  act  of  giving  food  is  regarded  by  the  natives  as 
in  some  sense  the  essential  characteristic  of  motherhood. 

The  food  so  given  is  a  free  gift  for  which  the  child  makes 
no  concrete  return,  but  the  mother  is  anxious  lest  the  child 
go  hungry  and  she  is  compensated  for  her  care  by  her  joy 
in  the  growth  of  the  child  and  her  pride  in  its  achievements. 

^  It  may  be  objected  that  the  suckling  of  the  child  is  a  "natural"  and 
not  a  "cultural"  act.  But  at  least  it  is  an  act  which  is  included  among 
the  norms  of  latmul  culture  and  which  has  no  doubt  been  culturally 
modified  in  many  ways.  We  are  justified  in  regarding  the  suckling  of  the 
child,  from  a  structural  point  of  view,  as  one  of  the  given  facts  of  latmul 
culture. 


76  Structural  Analysis  of  the 

The  tone  of  her  pride  is  well  illustrated  by  the  incident  in 
the  naven  in  which  the  homicide  marches  over  the  prostrate 
and  naked  bodies  of  his  "mothers"  and  they  say:  "That  so 
small  place  out  of  which  this  big  man  come ! "  The  attitude 
is  one  of  humbleness  and  abnegation  of  self  in  pride  and  joy 
at  the  achievement  of  the  oflPspring.  Lastly  we  have  the 
behaviour  of  the  mother  as  comforter  of  the  child ;  she  stops 
its  crying  and  stays  beside  it  in  sickness;  when  the  child  is 
too  sick  to  eat,  she  herself  will  go  without  food.^  I'he  position 
of  the  mother  may  thus  be  summed  up  under  headings  of: 
food-giving,  pride  and  self-abnegation,  comforting. 

These  aspects  of  the  mother's  position  are  clearly  stressed 
in  the  behaviour  of  the  wau.  We  see  him  as  food-giver: 
I  have  already  mentioned  that  there  is  a  continual  stream  of 
ceremonial  gifts  of  food  flowing  from  the  zvau  to  the  laua. 
At  small  naven  ceremonies,  a  fowl,  but  on  more  important 
occasions,  a  pig,  will  be  presented.  We  may  regard  these  gifts 
as  analogous  to  the  constant  feeding  of  the  child  by  its  mother. 
It  is  articulately  stated  by  the  natives  that  the  wau  is  anxious 
lest  his  laua  be  hungry.  A  small  boy  is  brought  up  to  believe 
that  unless  he  is  careful  how  he  asks  his  wau  even  for  a  single 
yam,  his  wau  will  exclaim  with  exaggerated  anxiety,  "What! 
Is  my  laua  hungry?"  and  jump  to  his  feet  and  run  to  kill 
his  pig  and  present  it  to  the  boy. 

We  see  the  wau  combining  self-abnegation  with  pride  in 
the  boy's  achievements:  The  attitude  of  vicarious  pride  is 
daily  exemplified  by  the  wau's  hailing  the  boy  with  a  series 
of  names  of  outstanding  totems  of  the  maternal  clan,  and 
more  dramatically,  whenever  the  laua  makes  a  conspicuous 
public  appearance,  e.g.  in  debate,  by  the  wau's  throwing  lime 
over  him.  Again  the  attitude  of  pride  is  dramatically  com- 
bined with  modesty  and  self-abnegation  in  the  navefi  cere- 
monies when  the  wau  wears  filthy  female  garments,  shaming 
himself  to  express  his  pride  in  his  laua.  Indeed,  even  at  this 
stage  of  our  enquiry  we  may  conclude  that  a  great  part  of 

^  I  have  no  evidence  that  this  detail  of  a  mother's  behaviour  is  carried 
over  into  the  behaviour  of  the  wan,  but  in  a  certain  stage  of  the  cere- 
monies of  initiation,  the  term  nyame  (mother)  is  applied  to  the  elder  age 
grade,  and  during  this  period  the  members  of  this  grade  observe  food 
taboos  with  the  novice. 


Wau-laua  Relationship  77 

the  wau's  behaviour  in  naven  is  an  exaggerated  statement  of 
this  particular  aspect  of  maternity,  the  mother's  self-abnega- 
tion combined  with  pride  in  the  boy.  But  many  details  of 
the  naven  still  remain  to  be  accounted  for — especially  the 
element  of  burlesque  in  the  wau's,  behaviour  and  the  waus 
offering  of  his  buttocks  to  laua. 

We  see  the  wau  as  a  comforter  in  the  ceremonies  of  initia- 
tion. At  this  time  he  is  called  "mother"  and  he  actually 
behaves  as  a  comforting  mother  to  the  little  boy  during  the 
grim  first  day  of  initiation  when  the  child  is  separated  from 
his  real  mother  and  undergoing  scarification.  The  waus  first 
act  in  initiation  is  to  lift  his  laua  pick-a-back — the  manner 
in  which  mothers  often  carry  their  children — and  carry  him 
to  the  upturned  canoe  on  which  the  scarification  is  to  be 
performed  (cf.  Plate  Vb). 

During  the  operation  the  zvau  sits  on  the  canoe,  and  the 
little  boy  sits  at  first  in  his  wau's  lap  while  the  incisions  are 
made  around  his  breasts.  After  that  the  boy  lies  with  his 
head  in  his  zvau's  lap  and  his  back  exposed  to  the  knives  of 
the  initiators.  The  cutting  of  the  back  is  much  more  painful 
than  the  preliminary  cutting  of  crescents  round  the  breasts 
and  the  little  boy  screams  and  clutches  as  hard  as  he  can 
with  his  arms  round  his  "mother's"  waist.  When  he  screams 
the  wau  will  say,  "Don't  cry,"  and  grasps  him,  holding  the 
boy  still  and  at  the  same  time  responding  to  his  frantic 
clutches.  When  the  cutting  is  at  last  finished,  the  wau  again 
takes  the  child  pick-a-back  and  carries  him  to  a  pond  where 
the  blood  is  washed  off.  Then  the  wau  carries  him  back  to 
the  ceremonial  house  where  he  soothes  the  boy's  cuts  with 
oil  gently  applied  with  a  feather  (cf.  Plates  X,  XI). 

Turning  now  to  the  reciprocal  behaviour  of  the  boy  towards 
his  wau  we  can  trace  again  points  in  which  it  is  modelled  on 
the  type  of  his  behaviour  to  his  mother.  In  the  attitude 
towards  the  mother  there  is  a  strong  emphasis  on  loyalty: 
I  have  already  published^  an  incident  in  which  a  son  inter- 
fered in  a  quarrel  between  his  mother  and  her  second  hus- 
band. The  husband  had  lost  his  temper  and  beaten  the  woman 
seriously  with  fire  sticks.  Her  son  entered  his  stepfather's 
^  Oceania,  1932,  p.  286. 


78  Structural  Analysis  of  the 

house  and  took  her  away  and  lodged  her  in  his  own  house 
for  several  days.  This  same  emphasis  on  loyalty  is  brought 
out  in  the  behaviour  towards  mother's  brother.  Whenever 
a  quarrel  starts  in  which  his  wau  is  involved  it  is  said  that 
the  boy  should  take  his  wau's,  side,  even  if  the  quarrel  is  one 
in  which  his  father  is  opposed  to  his  mother's  brother. 

The  same  loyalty  is  emphasised  in  matters  connected  with 
esoterica — spells,  mythology,  etc.  (cf.  p.  37). 

As  a  complement  to  the  wau's  pride  in  the  laua's  achieve- 
ments, boasting  is  developed  in  the  latter.  He  will  swagger 
in  his  ivau's  presence  and  will  say  with  great  gestures: 
"I  am  the  big  man  of  this  village!"  "I  am  the  house-post 
which  supports  this  village ! "  But  at  the  same  time  he  will 
not  swagger  overmuch,  and  especially  he  must  not  insult  his 
wau,  lest  the  latter  jump  up  and  rub  his  buttocks  on  the 
laua's  shin.  In  the  easier  relationship  between  a  boy  and  his 
mother,  this  boasting  behaviour  is  not  conspicuous,  and  its 
development  in  the  wau-laua  relationship  is  probably  corre- 
lated with  the  generally  exaggerated  character  of  the  wau's 
behaviour. 

In  this  exaggeration,  the  behaviour  of  the  wau  differs  from 
that  of  a  mother.  She  gives  food  easily  to  her  child,  but  when 
the  wau  gives  food  he  either  presents  pigs  in  a  ritual  context, 
or  jumps  up,  and,  with  a  histrionic  gesture,  kills  a  pig  in 
response  to  the  laua's  demand  for  a  yam.  The  mother  carried 
the  small  child  on  her  shoulders  because  that  was  a  con- 
venient way  of  carrying  it;  when  the  wau  carries  the  boy 
pick-a-back,  the  act  is  an  incident  in  the  ceremonial  of 
initiation  when  the  boy  is  grown  and  well  able  to  walk.  The 
mother  may  smile  easily  at  her  boy's  achievement,  but  the 
wau  is  impelled  at  least  to  throw  lime  over  him.  In  every 
instance  the  behaviour  of  the  wau  is  an  exaggerated  and 
dramatic  version  of  that  of  the  mother.  The  formulation  of 
the  social  identity  between  a  man  and  his  sister,  between  wau 
and  mother,  may  be  a  sufficient  explanation  or  description 
of  the  resemblances  in  the  behaviour  of  these  two  relatives, 
but  the  elements  of  burlesque  and  exaggeration  in  the  wau's 
behaviour  still  remain  to  be  explained. 

2.  We  may  now  consider  the  patterns  of  behaviour  typical 
of  the  relationship  between  a  man  and  his  sister's  husband 


Wau-laua  Relationship  79 

and  the  extent  to  which  these  patterns  are  adopted  by  the 
wau.  The  chief  characteristic  of  the  brother-in-law  relation- 
ship among  the  latmul  is  a  mutual  ambivalence  about  the 
fact  that  one  man  has  given  his  sister  to  be  the  wife  of  the 
other.  A  feeling  of  mutual^  indebtedness  and  distrust  is 
culturally  expected,  and  a  considerable  bride  price  is  given 
to  the  brothers  of  the  bride  by  the  husband.  Although  this 
transaction  is  phrased  as  waingga  (purchase),  a  considerable 
quantity  of  valuables  are  loaded  on  to  the  wife's  body  by  her 
brothers  when  she  goes  to  her  new  home  and  these  valuables 
are  a  return  presentation  to  the  husband.  The  natives  say 
quite  articulately  that  the  purpose  of  the  bride  price  is  "to 
prevent  the  wife's  relatives  from  using  sorcery  against  the 
husband",  and  they  say  that  this  resort  to  sorcery  will  be 
staved  off  only  for  a  limited  time  by  even  the  biggest  bride 
price.  The  sense  of  indebtedness  therefore  remains  and  the 
wife's  relatives  have  always  the  right  to  call  on  the  husband 
for  help  in  any  task,  like  house-building,  for  which  a  crowd 
of  manual  labourers  is  necessary.  When  the  task  is  completed 
the  wife's  people  will  stand  a  small  feast  for  the  labourers 
or  they  will  at  least  distribute  coconuts  to  them.  This  largess 
is  of  the  nature  of  a  complimentary  presentation  and  is 
usually  quite  inadequate  as  a  payment  for  work  done. 

But  the  brother-in-law  relationship  is  ambivalent;  it  is 
characterised  not  solely  by  indebtedness  and  distrust  but  also 
by  co-operation.  In  many  cases  the  bride's  brother  takes  an 
active  part  in  the  arrangements  for  the  marriage.  He  may 
assist  his  friend,  the  future  husband,  in  his  courtship  by 
putting  magic  in  his  sister's  fireplace  to  cause  her  to  fall  in 
love ;  and  finally  he  may  lead  her  by  night  from  their  house 
to  that  of  the  bridegroom.  On  arriving  there,  the  man  and 
his  sister  are  ceremonially  received  by  the  future  husband. 
The  girl's  brother  is  made  to  sit  on  a  stool  while  the  future 
husband  puts  oil  on  his  face,  cuts  his  hair  and  finally  puts 
valuables  on  his  shoulder.^  Next  morning  the  village  wakes 
to  find  the  marriage  a  fait  accompli. 

^  Cf.  the  quarrelling  between  wife's  fathers  and  husband's  fathers, 
described  on  p.  105 ;  Kwoshimba  versus  Kili-mali. 

*  Note  that  the  ceremonial  acts  performed  by  the  future  husband  are 
the  same  as  those  performed  by  an  offender  in  making  amends  to  the  man 
he  has  offended  (cf.  p.  98). 


8o  Structural  Analysis  of  the 

In  many  cases  too  the  opposition  between  brothers-in-law 
is  treated  lightly  and  used  as  a  subject  for  joking.  If  a  man 
is  idly  watching  his  tawontu  doing  some  job,  the  latter  will 
say  merrily  enough,  "Watching  your  tawontu}  1  am  not  that 
sort  of  man.  You  should  see  me  help  my  tawontus."  It  is 
probable  too  that  the  opposition  between  brothers-in-law  is 
reduced  to  a  minimum  in  those  cases  in  which  the  relation- 
ship between  them  is  symmetrical,  based  upon  an  exchange 
of  sisters.  But  no  native  statement  was  collected  to  this 
effect. 

Even  without  this  exchange  there  is  an  insistence  that  a 
man  should  be  loyal  to  his  wife's  brother — a  feature  in  which 
this  relationship  resembles  that  with  the  mother  and  the 
mother's  clan.  Indeed,  we  find  here  the  same  type  of  precept 
which  we  noted  in  the  relationship  to  the  mother's  clan — in 
this  case,  that  a  man  should  support  his  wife's  people  in  all 
their  quarrels,  even  against  his  own  clan.  In  this  respect 
then,  father  and  son  are  alike  in  owing  absolute  allegiance 
to  the  same  clan,  and  the  insistence  upon  loyalty  to  the  wau 
may  be  phrased  as  consistent  both  with  identification  of  that 
relative  with  the  boy's  mother  and  with  the  identification  of 
the  boy  with  his  father. 

When  a  quarrel  arises  between  a  man's  maternal  clan  and 
the  clan  into  which  he  has  married,  he  should  either  stand 
up  between  the  disputants  and  stop  the  fight,  or  hold  his 
peace.  In  general,  these  precepts  are  to  be  regarded  as  state- 
ments of  what  a  man  ought  to  do  rather  than  as  statements 
of  what  is  usually  done.  But  I  have  on  several  occasions  seen 
men  intervene  in  brawls  in  the  ceremonial  house.  Unfor- 
tunately I  never  examined  the  kinship  ties  between  them  and 
the  disputants.  At  such  times  I  was  myself  anxious  lest  some 
serious  damage  might  be  done  which  would  call  down  the 
wrath  of  the  Government  upon  my  village  and  perhaps 
deprive  me  of  the  services  of  valuable  informants.  I  pre- 
sumed that  the  motives  of  the  peacemakers  were  a  product 
of  modern  culture  contact  and  fear  of  the  Government,  It 
never  occurred  to  me,  even  though  I  knew  the  precepts,  that 
such  kill-joy  conduct  could  be  culturally  normal  to  the  latmul. 

Besides  the  insistence  upon  loyalty,  other  aspects  of  the 


Wau-laua  Relationship  8i 

brother-in-law  relationship  are  taken  over  into  the  wau-laua 
relationship.  A  husband  gives  valuables  and  labour  and  re- 
ceives a  wife — and  these  gifts  are  tied  together  by  a  sense  of 
indebtedness.  This  uneasy  economic  aspect  of  the  brother- 
in-law  relationship  has  been  taken  over  into  the  pattern  of 
the  wau-laua  relationship.  We  saw  above  that  the  wau  makes 
considerable  presentations  of  food — especially  pigs  and  fowls 
— to  the  sister's  son.  But  these  gifts — unlike  the  corresponding 
food  gifts  of  the  mother — prompt  the  boy  to  a  sense  of 
indebtedness,  and  he  makes  in  return  presentations  of  shell 
valuables  to  his  wau.  Actually  in  many  cases  the  boy  has  no 
wealth  of  his  own  with  which  to  recompense  his  wau  and  it 
is  the  father  of  the  boy  who  provides  the  shells  for  the  pre- 
sentation, a  fact  which  illustrates  the  identification  of  sister's 
husband  with  sister's  child  most  vividly  in  the  present  con- 
text. Stated  from  the  wau's  point  of  view :  he  presents  a  pig 
to  his  sister's  son  and  his  sister's  son  and  his  sister's  husband 
combine  their  wealth  to  make  a  return  present  of  valuables. 
Lastly  we  may  consider  the  extraordinary  conduct  of  the 
wau  in  presenting  his  buttocks  to  his  laua.  Such  conduct  is 
of  course  not  characteristic  of  a  mother,  but  I  have  a  casual 
mention  in  mythology  of  a  man  who  rubbed  his  buttocks 
on  the  leg  of  the  man  who  was  marrying  his  sister.  If  we 
bear  in  mind  the  identification  of  a  man  with  his  sister,  this 
conduct  is  comprehensible — at  least  from  a  structural  point 
of  view.  The  man  expressed  his  relationship  to  his  sister's 
husband  by  ritually  making  a  sexual  gesture  in  which  he 
identified  himself  with  his  sister.  In  the  case  of  the  wau,  the 
position  is  more  complicated,  but  his  behaviour  becomes 
perfectly  logical  if  we  think  of  the  two  identifications  upon 
which  it  is  based.  For  the  sake  of  clarity  we  may  state  these 
identifications  as  if  the  mother's  brother  were  speaking,  viz. 
" I  am  my  sister "  and  "My  nephew  is  my  sister's  husband." 
If  now  we  consider  these  two  identifications  simultaneously 
it  is  perfectly  "logical"  for  the  wau  to  offer  himself  sexually 
to  the  boy,^  because  he  is  the  boy's  wife. 

^  I  should  perhaps  stress  the  fact  that  this  interpretation  of  the  wau's 
gesture  is  my  own  and  is  not  based  upon  native  statements.  It  is  con- 
ceivable that  the  gesture  may  be  symbolic  of  the  birth  of  laua  from  wau, 
a  further  stressing  of  the  maternal  aspects  of  the  wau's  position. 


82  Structural  Analysis  of  the 

This  last  piece  of  analysis  may  seem  to  be  fanciful,  but  the 
evidence  goes  to  show  that  there  is  some  such  process  of 
logical  symbolism  underlying  this  piece  of  ritual.  Upon  this 
hypothesis  the  zvau's  exclamation :  "  Lan  men  to!  "  "  Husband 
thou  indeed!"^  becomes  immediately  understandable.  Lastly 
the  fact  that  the  wau^s  sexual  gesture  is  said  to  act  as  a  strong 
stimulus  to  make  the  laua  run  to  get  valuables  in  order  to 
present  them  to  his  wau  is  explained  if  this  presentation  is 
really  in  some  sort  a  bride  price. 

But  although  the  structural  position  of  the  wau  is  such  that 
his  actions  in  the  naven  ceremony  can  be  described  as  its 
logical  development,  we  still  do  not  know  why  the  culture 
should  have  followed  up  this  line  of  logic.  Why  should  the 
culture  have  followed  these  various  identifications  to  their 
ultimate  reductio  ad  absurdum,  a  wau  offering  himself  sexually 
to  his  laua}  Such  formulated  identifications  between  a  man 
and  his  sister  and  between  a  man  and  his  father  can  be  traced 
in  many  primitive  communities,  but  even  the  simpler  ele- 
ments of  the  naven  ritual — the  wau  dressing  as  a  woman, 
etc. — are  comparatively  rare  in  such  societies;  and  such  a 
phenomenon  as  the  wau's  sexual  gesture  is  perhaps  only  to 
be  found  in  the  Sepik  area.  We  must  therefore  conclude  that 
while  the  structural  position  lays  down  possible  lines  along 
which  the  culture  may  develop,  the  existence  of  such  lines 
does  not  explain  why  the  culture  should  select  them  for 
emphasis.  We  have  demonstrated  the  existence  of  the  lines 
but  we  have  still  to  demonstrate  the  "motive  force"  which 
has  caused  the  culture  to  run  along  them. 

3.  Lastly  we  have  to  consider  the  various  points  in  the 
waw's  behaviour  which  cannot  be  described  by  saying  either 
that  the  wau  is  a  male  mother  of  the  laua,  or  that  the  laua  is 
the  sister's  husband  of  the  wau. 


^  The  evidence  from  this  exclamation  is,  however,  not  as  strong  as  its 
hteral  translation  would  indicate,  since  the  words,  lan  men  to,  are  also 
constantly  used  as  an  expression  of  submissiveness,  a  demand  for  com- 
passion and  a  cry  of  surrender.  The  phrase  can  even  be  used  in  addressing 
a  female  oppressor  and  is  then  modified  to  "lain  nyin  to".  (Nyin 
feminine,  second  person,  singular  pronoun.  The  modification  from  lan 
to  lain  is,  I  think,  not  expressive  of  change  of  gender  but  merely 
"euphonic".) 


Wau-laua  Relationship  83 

{a)  The  wau  will  not  take  fire  from  the  laud's  house.  If 
the  fire  goes  out  in  his  own  house,  the  wau  will,  it  is  said, 
go  and  light  his  cigarette  at  the  laua's  fire  and  from  the 
cigarette  will  light  his  own  fire,  but  he  will  not  take  a  burning 
fire  stick  from  the  laua's  fire.  This  type  of  avoidance  is  more 
typical  of  the  boy's  father,  who,  as  I  have  mentioned,  will 
never  under  any  circumstances  eat  of  food  collected  by  his 
son — fire  and  food  being  closely  analogous  in  latmul  thought. 

It  would  appear  that  the  wau's  avoidance  of  taking  fire 
from  his  laud's  house  is  perhaps  structurally  described  by  an 
identification  of  the  wau  with  the  lauds  father;  and  this  is 
the  more  probable  since  the  complete  series  of  taboos  on 
eating  food  collected  by  the  boy  is  applied  to  the  father's 
tamhinien.  The  tambinien  is  a  partner  and  the  behaviour 
between  tatnbiniens  is  modelled  upon  the  same  general  lines 
as  the  reciprocal  co-operation  between  men  who  have  ex- 
changed sisters.  Thus  it  is  not  surprising  to  find  that  a 
similar  identification  obtains  to  some  extent  between  simple 
brothers-in-law. 

{b)  The  wau  resembles  a  father  in  another  point;  like  a 
father  he  avoids  entering  upon  crudely  commercial  trans- 
actions with  his  laua.  But  it  is  not  perfectly  clear  that  this 
resemblance  between  the  two  relatives  is  due  to  an  identifica- 
tion between  them  and  not  to  some  coincidence  whereby 
diflFerent  causes  operate  upon  the  two  relatives  to  induce 
superficially  similar  behaviour.  In  the  case  of  a  father,  the 
avoidance  of  commercial  relations  with  the  son  is,  no  doubt, 
a  part  of  the  general  avoidance  of  profiting  at  the  son's 
expense.  In  the  case  of  the  wau,  I  was  told  that  the  avoidance 
is  specifically  connected  with  the  fact  that  this  relative  has 
received  a  part  of  the  bride  price  paid  for  the  lauds  mother ; 
but  I  do  not  know  why  this  fact  should  make  it  uncomfortable 
or  unfitting  for  a  wau  to  trade  with  his  laua ;  nor  do  I  know 
whether  commercial  transactions  occur  between  brothers- 
in-law, 

{c)  Another  usage  may  doubtfully  be  referred  to  the  same 
identification :  when  a  laua  performs  some  service  for  his  wau 
he  is  spoken  of  as  the  latter's  "  dog  ".  I  do  not  know  exactly 
what  details  of  the  relationship  between  wau  and  laua  are 


84  Structural  Analysis  of  the 

here  equated  with  what  details  of  the  relationship  between 
a  man  and  his  dog ;  but  it  is  worth  remarking  that  the  owner 
of  a  dog  is  ordinarily  spoken  of  as  its  "father". 

{d)  I  have  already  described  the  series  of  details  of  be- 
haviour which  seem  to  equate  the  laua  with  the  ancestors 
of  the  maternal  clan  and  suggested  that  this  equation  is  based 
on  a  similar  orientation  of  pride  towards  ancestors,  lauas  and 
achievement. 

{e)  The  behaviour  of  the  wau  is  exaggerated.  It  is  not  a 
mere  copy  of  that  of  a  mother  or  brother-in-law,  but  is  a 
sort  of  burlesque  of  the  behaviour  of  these  relatives.  This 
element  of  burlesque  must  for  the  present  be  described  as 
peculiar  to  the  wau  relationship  and  not  as  derived  from  any 
pattern  of  identifications. 

Ill 


From  the  foregoing  analysis  it  follows  that  we  may  sum  up 
the  principal  characteristics  of  the  wau'?.  position  by  means 
of  a  diagram  (Fig.  i).  In  this  diagram  the  genealogical  position 
of  wau  and  laua  is  shown  and  three  arrows  ^re  drawn  repre- 
senting the  three  identifications  on  which  the  behaviour  of 
wau  and  laua  is  based.  Arrow  I  marks  the  identification  of 
wau  with  his  sister,  according  to  which  wau  will  behave  as 
a  mother,  and  the  laua  will  behave  as  a  son  towards  a  mother. 
Arrow  II  marks  the  identification  of  the  laua  with  his  father 
according  to  which  wau  will  treat  laua  as  if  he  were  a  sister's 
husband  and  laua  will  treat  wau  as  if  he  were  a  wife's  brother. 
The  combination  of  identifications  I  and  II  is  the  basis  for 
the  wau's  treatment  of  laua  as  if  he  were  a  wife. 

Arrow  III  marks  the  identification  between  wau  and  laua's 


Wau-laua  Relationship  85 

father,  according  to  which  wau  will  behave  as  if  he  were 
laua's  father,  and  laua  will  behave  as  if  he  were  waus  son. 
Since  this  last  identification  is  much  more  problematical  than 
the  other  two  and  since  I  have  no  record  of  any  behaviour 
of  the  laua,  which  would  fit  this  pattern,  I  have  indicated 
this  identification  with  only  a  dotted  line  on  the  diagram. 

Our  experiment  has  shown  that  almost  the  whole  of  the 
wau's  cultural  behaviour  can  be  analysed  in  terms  of  two 
identifications.  Of  the  details  which  remain  unaccounted 
for,  some  can  be  referred  to  a  third  identification  but  a 
few  still  defy  analysis  in  these  terms.  We  still  do  not  know 
why  the  wau's  behaviour  tends  to  be  exaggerated  and  comical, 
why  the  laua  is  regarded  as  the  ancestors  of  the  maternal 
clan,  and  why  this  culture  has  followed  the  logic  of  its 
identifications  to  such  extreme  conclusions.  These  problems 
must  be  left  for  solution  in  terms  of  aspects  of  latmul  culture 
other  than  the  purely  structural. 


CHAPTER  VII 

The  Sociology  of  Naven 

HITHERTO  we  have  confined  ourselves  rigidly  to  what 
I  have  called  the  Structural  point  of  view.  We  have 
put  ourselves  in  the  place  of  a  hypothetical  intellect 
inside  the  culture  and  have  shown  that  to  such  an  intellect 
the  behaviour  of  wau  and  laua  in  the  naven  is  in  some  sense 
logically  consistent  with  the  other  given  facts  of  the  latmul 
culture.  But  in  the  present  section  we  shall  consider  the 
naven  from  an  entirely  different  point  of  view,  that  of  an 
observer  outside  the  culture  and  interested  in  the  integration 
and  disintegration  of  latmul  communities.  We  shall  consider 
only  those  aspects  of  naven  which  are  relevant  to  the  well- 
being  of  that  abstract  unity,  the  Society  as  a  whole.  This  shift 
of  viewpoint  involves  a  complete  re-orientation  towards  the 
culture  and  a  complete  re-phrasing  of  our  problems. 

The  fundamental  problem  with  which  we  are  now  con- 
cerned is :  Does  the  naven  have  any  effects  on  the  integration 
of  Society?  This  problem  I  shall  modify  by  making  an 
assumption  of  a  type  not  unfamiliar  to  anthropologists.  It  is 
evident  that  the  naven  ceremony  is  an  expression  and  a 
stressing  of  the  kinship  link  between  the  wau  and  laua  con- 
cerned, and  from  this  I  shall  assume  that  by  this  stressing 
the  link  is  strengthened. ^  In  the  light  of  this  assumption, 
the  problem  of  the  sociological  functions  of  naven  becomes : 
What  effects  has  the  strengthening  of  this  link  upon  the 
integration  of  the  Society  as  a  whole  ? 

To  answer  this  question  we  shall  have  to  ask  many  sub- 
sidiary questions,  and  the  form  of  these  will  contrast  sharply 
with  the  form  of  the  subsidiary  questions  which  we  asked 
in  our  structural  enquiry. 

'^  This  is  not  a  sociological  assumption  but  refers  to  a  process  of  psycho- 
logical conditioning  of  the  individuals  concerned.  The  question  of  its 
validity  should  therefore  be  examined  not  in  the  present  chapter  but  in 
a  chapter  devoted  to  the  study  of  character  formation  among  the  latmul, 
a  subject  which  I  did  not  investigate  in  the  field. 


The  Sociology  of  Naven  87 

At  the  outset  it  appears  that  a  sociological  enquiry  should 
be  documented  with  statistics,  a  method  of  treatment  which 
would  have  been  excessively  awkward  in  the  study  of  cultural 
structure.  At  the  most,  the  only  statistical  questions  which 
could  have  been  relevant  to  our  structural  enquiry  would 
have  taken  the  form :  "  How  many  details  of  behaviour  imply 
so-and-so?"  But  in  the  present  section,  the  whole  phrasing 
of  the  problem  being  different,  the  type  of  statistical  question 
which  we  shall  ask  will  take  the  general  form:  "How  many 
individuals  do  so-and-so?"  a  form  of  question  to  which 
answers  could  have  been  collected  in  the  field. ^ 

If  we  consider  a  simple  example  from  our  own  culture  this 
difference  in  the  framing  of  the  statistical  questions  will  be 
clear:  Fathers  beat  their  children.  This  fact  could  be  struc- 
turally investigated  and  it  could  be  shown  that  the  beating 
of  children  was  consistent  with  other  details  of  the  father- 
child  relationship;  and  this  consistency  would  be  equally 
demonstrable  though  our  note-books  only  contained  re- 
ference to  one  case  of  child-beating.  But  if  we  are  investi- 
gating the  same  fact  from  a  sociological  point  of  view  we  shall 
need  statistics.  We  may  guess  that  the  beating  of  children 
is  important  for  the  general  maintenance  of  order  in  Society. 
But  to  demonstrate  this  function  conclusively  we  need  sta- 
tistics which  will  answer  questions  of  the  general  type,  "What 
percentage  of  such  and  such  individuals  do  so-and-so?" 
Specifically  in  the  present  instance  we  shall  want  to  know : 

1.  What  percentage  of  fathers  beat  their  children? 

2.  What  percentage  of  beaten  children  in  later  life  go  to 
prison  for  riotous  behaviour? 

3.  What   percentage   of  unbeaten   children   later  go   to 
prison?  Etc.  etc. 

From  this  example  we  may  deduce  that  the  emphasis  of 
the  sociological  questions  is  likely  to  be,  not  "How  many 
details  of  cultural  behaviour. ,  .  ?"  but  rather  "How  many 
individuals . . .  ? "    Further,  it  appears  that  we  must  expect 

^  Unfortunately  I  have  no  statistics  and  took  no  random  samples.  The 
conclusions  of  this  chapter  are  therefore  unproven.  The  chapter  itself 
is  only  included  for  the  sake  of  giving  an  illustration  of  the  problems 
and  methods  of  approach  of  sociology  in  the  strict  sense  of  the  word. 


88  The  Sociology  of  Naven 

in  a  sociological  enquiry  to  be  concerned  with  classes  of  indi- 
viduals— "beaten  children",  "unbeaten  children",  "fathers 
who  beat",  etc.  In  fact,  it  will  be  our  business  to  classify, 
not  details  of  behaviour  according  to  their  implicit  assump- 
tions, but  individuals  according  to  their  behaviour. 

Returning  to  the  problem  of  nave^i  which  I  stated  above  in 
terms  of  the  sociological  importance  of  the  link  between  wau 
and  laua,  one  fact  now  stands  out  as  likely  to  be  a  clue  to  the 
position :  it  is  the  classificatory  waus  who  perform  naven.^ 

The  relevance  of  this  fact  becomes  clear  when  we  examine 
the  marriage  system  of  the  latmul  and  the  patterns  of  be- 
haviour between  groups  of  individuals  who  are  linked  together 
by  marciages  in  past  generations. 

In  many  primitive  societies  we  find  marriage  regulated  in 
a  positive  manner  by  kinship.  Not  only  are  there  prohibitions 
of  marriage  with  certain  relatives,  but  also  positive  injunc- 
tions which  constrain  a  man  to  choose  as  wives  women  who 
are  related  to  him  in  some  specific  manner,  real  or  classifica- 
tory. Under  such  systems  the  links  of  affinal  allegiance  and 
indebtedness  are  regularly  renewed  in  succeeding  generations. 
But  in  latmul  society  this  is  not  the  case.  (Lacking  statistics 
I  must  here  substitute  a  description  of  the  marriage  system 
in  formal  terms.)  The  culture  does,  it  is  true,  contain  a  great 
many  formulations  which  would  regulate  marriage  in  a  posi- 
tive manner  if  they  were  consistently  obeyed.  We  find  for 
example  such  statements  as : 

I .  "A  woman  should  climb  the  same  ladder  that  her  father's 
father's  sister  climbed",  i.e.  she  should  enter,  as  a  bride,  the 
house  which  her  father's  father's  sister  entered.  This  is  a  way 
of  stating  that  a  woman  should  marry  her  father's  father's 
sister's  son's  son  (or  reciprocally  that  a  man  should  marry 

^  This  is  certainly  the  case  in  Palimbai  and  Kankanamun.  But  the 
position  in  Mindimbit  is  probably  rather  different.  In  the  only  naven 
which  I  studied  in  detail  in  the  latter  place,  the  naven  for  a  girl  who  had 
caught  a  fish,  it  was  the  own  waus  who  took  part.  In  Palimbai  the  naven 
for  a  boy  who  had  made  a  new  canoe  was  performed  by  classificatory 
waus  and  it  was  in  Palimbai  that  I  was  told  that  a  man  has  two  sorts  of 
tawontu,  those  who  receive  a  share  of  the  bride  price  of  the  man's  wife 
and  those  who  will  celebrate  naven  for  the  man's  children.  It  is  possible 
that  the  sociological  functions  of  naven  in  Mindimbit  differ  profoundly 
from  those  I  have  suggested  for  Palimbai  and  Kankanamun. 


The  Sociology  of  Naven  89 

his  father's  mother's  brother's  son's  daughter,  or  iai.^  Such 
a  marriage  rule  if  carried  out  consistently  would  lead  to  a 
repetitive  system  and  the  affinal  links  would  then  be  renewed 
in  every  alternate  generation.  But  the  rule  is  not  so  carried 
out  and  we  find  other  conflicting  formulations. 

2.  "The  daughter  goes  as  payment  for  the  mother."  This 
is  a  way  of  stating  that  a  man  marries  his  father's  sister's 
daughter ;  and  this  practice  again  would  lead  if  it  were  con- 
sistently obeyed  to  a  repetitive  system,^  and  one  which, 
though  conflicting  with  the  iai  marriage  mentioned  above, 
is  still  generically  connected  with  it  in  a  curious  way. 


This  generic  relationship  between  the  two  systems  is  illus- 
trated in  the  diagram,  Fig.  2.  The  marriage  between  A'^ 

^  Cf.  Oceania,  1932,  p.  263.  There  is  a  second  cliche  which  defines 
iai  marriage:  " Laua's  son  will  marry  wau's  daughter."  This  refers  to 
marriage  of  a  man  with  his  father's  mother's  brother's  daughter,  a  relative 
who  is  also  called  iai.  This  cliche  was  collected  both  in  Mindimbit  and 
in  Palimbai,  while  the  clichd  referring  to  the  ladder  was  collected  only 
in  Kankanamun.  It  is  possible  that  in  Mindimbit  the  sociological  function 
of  naven  is  the  promotion  of  marriage  between  laua's  son  and  wau's 
daughter.  Thus  the  naven  becomes  a  "device"  not  for  perpetuating 
affinal  links  which  would  not  otherwise  be  repeated  in  the  community, 
but  for  bringing  about  their  repetition.  I  have,  however,  no  statistics  or 
native  statements  which  would  support  this  theory. 

*  Cf.  Dr  Fortvme,  on  Kinship  Structure,  Oceania,  1933,  pp.  1-9. 
Among  the  latmul,  however,  this  type  of  exchange  is  not  strictly 
confined  to  the  regulation  of  marriage.  In  certain  cases,  when  the  off- 
spring of  the  marriage  are  male  instead  of  female,  one  of  the  sons  will 
be  sent,  while  he  is  still  a  baby,  for  adoption  into  the  family  and  clan  of 
the  man  who  gave  his  sister  to  the  father  for  wife,  i.e.  the  boy  is  adopted 
by  his  mother's  brother,  to  whom  he  goes  as  payment  for  his  mother. 


90  The  Sociology  of  Naven 

and  6'$  is  an  iai  marriage  which  repeats  the  original  marriage 
(^  (^  =  6  ?)  in  the  grandparent  generation.  The  difference 
between  the  iai  system  and  the  father's  sister's  daughter 
system  turns  on  whether  the  marriage,  jc  $  =  Y^,  occurs  in 
the  intervening  middle  generation.  If  this  marriage  occurs, 
the  whole  system  is  converted  into  one  based  upon  marriage 
with  the  father's  sister's  daughter,  and  the  final  marriage, 
A'(^  =  b'^,  becomes  a  marriage  of  a  man  with  his  father's 
sister's  daughter.  The  introduction  of  this  intervening  mar- 
riage is  linked  with  a  new  formulation: 

3.  "Women  should  be  exchanged."  This  is  more  usually 
stated  by  the  natives  in  reference  to  sister  exchange — a  man 
giving  his  sister  as  wife  to  the  man  whose  sister  he  himself 
marries.  Such  a  formulation,  stated  in  this  way  in  terms  of 
sister  exchange,  of  course  conflicts  with  either  iai  marriage 
or  father's  sister's  daughter  marriage.  But  the  fundamental 
concept  of  the  exchange  of  women  is  common  to  both  sister 
exchange  and  marriage  with  father's  sister's  daughter. ^ 

Thus  latmul  society  is  built  upon  three  formulations  in 
regard  to  marriage  which  though  they  conflict  are  still  inter- 
related in  a  curious  logical  manner.  The  father's  sister's 
daughter  marriage  is  comparable  with  the  iai  marriage  since 
both  fit  into  the  same  pattern  of  repetitions  in  alternate 
generations,  and  the  sister  exchange  is  comparable  with 
father's  sister's  daughter  marriage  since  both  depend  upon 
the  exchange  of  women. 

This  discussion  of  the  marriage  rules  was  introduced  to 
show  that  the  society  is  without  any  regular  repetitive  system 
whereby  the  affinal  links  might  be  regularly  renewed.  I  have 
elaborated  the  description  somewhat  more  than  is  necessary 
in  the  present  context  because  the  cultural  details  here  given 

^  It  is  quite  likely  that  the  marriage  with  father's  sister's  daughter  has 
been  evolved  in  the  culture  through  interaction  of  the  two  systems,  iai 
marriage  and  sister  exchange.  The  whole  kinship  vocabulary  would  point 
to  the  iai  system  as  the  older,  and  the  concept  of  exchange  of  women  may 
well  have  been  adopted  from  neighbouring  peoples.  In  this  connection 
it  is  interesting  that  one  of  the  young  men  whom  I  used  as  an  informant 
had  a  strong  impression  that  the  correct  term  for  father's  sister's  daughter 
was  iai.  But  after  we  had  discussed  the  matter  in  detail,  he  was  uncertain 
and  consulted  one  of  the  older  men  who  stated  definitely  that  na  was  the 
correct  term  for  this  relative. 


The  Sociology  of  Naven  91 

will  also  be  relevant  in  a  later  chapter.  We  shall  see  that  this 
society  constantly  dovetails  together  ideas  which  are  incom- 
patible and  that  there  is  perhaps  something  in  common 
between  the  muddled  logic  which  underlies  the  naven  ritual 
and  that  on  which  the  rules  of  marriage  are  based.  In  the 
discussion  of  the  totemic  system,  we  shall  see  (p.  128)  the 
capacity  of  the  natives  to  take  a  pride  in  the  schematic  quality 
of  their  system,  although  actually  that  system  is  riddled  with 
incompatibilities  and  fraud.  In  their  marriage  rules  too  the 
natives  take  a  pride  and  they  look  down  on  their  neighbours 
as  "dogs  and  pigs  who  mate  at  random". 

This  gibe  is  singularly  inappropriate  in  the  mouths  of  the 
latmul,  since  not  only  have  they  three  positive  marriage  rules 
which  conflict  one  with  another,  a  circumstance  which  would 
in  itself  introduce  a  random  element  into  their  mating,  but 
the  people  do  not  adhere  even  to  their  negative  rules.  These 
negative  rules  are  very  vague,  but  there  is  a  strong  feeling 
against  marrying  own  sister  and  I  never  collected  a  case  of 
marriage  of  this  kind.  The  next  strongest  ban  is  that  upon 
marriage  with  any  of  the  relatives  called  naisagut  (e.g.  wife's 
mother),  and  I  found  a  case  in  which  a  man^  had  sufficient 
standing  in  the  community  to  marry  his  own  wife's  own 
mother  and  this  while  the  wife  was  still  alive  and  married 
to  him.  He  was  a  great  sorcerer  and  at  the  same  time  a  great 
debater  and  fighter.  It  was  nobody's  business  to  say  him 
nay  in  this  individualistic  culture.  Such  marriages  are  rare — 
even  among  the  latmul — but  when  we  turn  to  the  less  strin- 
gent prohibitions,  as  of  marriage  with  classificatory  "  sisters  ", 
women  of  own  clan  with  whom  genealogical  connections  can 
easily  be  traced — we  find  that  such  women  are  sometimes  taken 
as  wives.  When  I  enquired  into  these  marriages,  I  was  generally 
answered  with  a  cliche :  "  She  is  a  fine  woman  so  they  married 
her  inside  the  clan  lest  some  other  clan  take  her,"  The  latmul 
even  rather  approve  of  these  endogamous  marriages  and  say 
that  from  endogamy  are  produced  long  and  widely  rami- 
fying lines  of  descendants — a  belief,  the  sociological  and 
cultural  significance  of  which  is  not  at  all  clear  to  me. 

Occasionally,    however,    the   members    of   a   large   clan, 

^  This  was  Mali-kindjin,  of  whose  character  a  sketch  is  given  on  p.  164. 


92  The  Sociology  of  Naven 

apparently  realising  that  they  are  becoming  too  endogamous, 
proceed  to  formulate  a  statement  that  one  half  of  the  clan 
belongs  to  the  bows  of  the  clan  war  canoe  and  the  other  half 
belongs  to  the  stem.  Henceforth  these  two  groups  are  re- 
garded as  separate  clans  for  purposes  of  marriage  and  what 
is,  in  a  sense,  really  endogamy  is  satisfactorily  phrased  as 
exogamy. 

Marriages  with  classificatory  "mothers"  are  also  not  un- 
common. 

Lastly,  besides  these  endogamous  marriages,  there  are 
many  marriages  with  outside  groups — with  women  captured 
in  war,  or  sent  as  peace  offerings,  women  met  on  trading 
expeditions,  etc.  So  that  we  may  sum  up  the  marriage  system 
by  saying  that  in  practice  marriage  occurs  very  nearly  at 
random;  and  when  we  consider  that  the  villages  are  very 
large,  with  a  population  ranging  from  two  hundred  to  a 
thousand,  it  seems  unlikely  that  an  important  affinal  link  will 
be  perpetuated  by  analogous  marriages  in  future  generations. 
If  therefore  these  old  affinal  links  are  necessary  for  the 
integration  of  the  community,  some  means  must  be  found  of 
diagrammatically  stressing  them,  a  function  performed  by 
naven. 

It  remains  for  us  to  consider  whether  these  linkings  are 
really  important  and  how  they  are  used  in  the  integration  of 
society.  In  this  connection  the  most  significant  fact  is  that 
the  formulations  of  proper  behaviour  towards  affinal  relatives 
can  be  applied  in  this  culture  not  only  to  own  wife's  own 
relatives  and  to  own  sister's  husband's  relatives,  but  can  be 
extended  to  a  whole  series  of  relatives  who  are  grouped  around 
this  central  nucleus  in  a  classificatory  manner.  Such  funda- 
mental principles  as  the  identification  of  siblings,  the  identi- 
fication of  a  man  with  his  father  and  with  his  father's  father, 
and  the  grouping  of  whole  clans  as  single  units,  are  applied 
in  this  extension  of  affinal  relationships.  But  it  would  not  be 
true  to  say  that  the  affinal  linkages  due  to  every  marriage 
are  extended  indiscriminately  in  every  direction  according  to 
this  scheme  of  identifications. 

The  formulations  of  proper  behaviour  contain  such  pre- 
cepts as  that  a  man  should  be  loyal  to  his  wife's  relatives  and 


The  Sociology  of  Naven  93 

should  help  them  in  the  work  of  building  their  houses  and 
their  ceremonial  houses ;  he  should  help  them  pull  their  canoe 
logs,  clear  their  gardens,  and  he  should  come  to  their  assist- 
ance in  all  their  quarrels.  If  all  affinal  linkages  were  observed 
no  one  would  be  able  to  quarrel  with  anyone  else  inside  the 
community,  and  everybody  would  have  to  go  everywhere 
and  do  everything  with  everybody  else,  since  the  genealogical 
links  are  actually  ubiquitous. 

One  need  only  spend  a  very  few  days  in  a  latmul  village 
to  realise  that  there  is  no  such  thorough  integration  of  the 
population  as  this  complete  and  indiscriminate  extension  of 
affinal  ties  and  duties  would  imply.  Quarrels  are  not  rare 
and  whenever  a  large  group  of  workers  is  required  for  some 
great  labour  such  as  house-building  or  repairing  the  cere- 
monial house,  it  is  difficult  to  get  the  men  together  and  diffi- 
cult to  get  them  to  co-operate.  But  in  the  end  the  groups 
can  be  got  to  work  and  such  co-operation  as  is  achieved  is 
certainly  made  possible  by  the  insistence  upon  affinal  ties 
which  extend  in  a  perfectly  definite  manner  along  certain 
chosen  lines  of  the  classificatory  system. 

There  are  two  main  types  of  extension  of  the  affinal 
system  and  various  collective  terms  are  applied  to  the  ex- 
tended groups  so  formed.  The  two  commonest  terms  are 
lanoa  nampa  (literally  "husband  people")  and  laua  nyanggu 
(" sister's-child  children" — the  word  "children"  in  this  con- 
text being  no  more  than  a  synonym  for  "people"). 

The  term  lanoa  nampa  is  strictly  most  applicable  to  the 
group  formed  by  the  first  type  of  extension  of  the  affinal 
system.  This  is  the  group  of  persons  whose  affinal  relation- 
ship results  from  some  definite  contemporary  marriage,  and 
it  contains  such  relatives  as :  sister's  husband  (lan-ndo),  sister's 
husband's  brother  (lan-ndo),  sister's  husband's  father  (laua) 
and  sister's  husband's  son  (laua).  Thus,  when  the  relation- 
ship depends  upon  a  recent  marriage,  the  term  Ian  is  extended 
to  include  various  lauas.  (The  reciprocal  collective  term  for 
relatives  of  the  wife  is  towa-naisagut.) 

In  the  second  type  of  extension  of  the  affinal  system,  the 
relatedness  of  the  groups  is  traceable  to  a  past  marriage,  but 
the  reality  of  the  connection  between  the  groups  is  demon- 


94  The  Sociology  of  Naven 

strated  by  the  naven  system.  In  this  type  of  extension,  the 
more  correct  term  for  the  group,  which  contains  the  relatives 
of  the  former  husband,  is  laua  nyanggu  (though  the  term 
lanoa  nampa  is  sometimes  used).  Reciprocally,  the  group 
containing  the  relatives  of  the  former  wife  is  called  zvau  nyame 
nampa.  Thus  the  term  laua  is  here  extended  to  include  classi- 
ficatory  sister's  husbands,  the  reverse  of  the  usage  mentioned 
above  in  which  the  term  Ian  is  extended  to  include  sister's 
children. 


A 


B 


cf  =  $  9  =  cd* 


^=*$      H  ^cf      ^?=^c? 


E  cT  {lam  of  Cd)  "^  D<^  {laua  of  b6) 

Figure  3. 

The  natives  are  perfectly  conscious  of  the  genealogical 
process  upon  which  the  formation  of  this  laua  group  of 
relatives  depends,  namely,  the  perpetuation  of  past  marriages. 
There  is  a  native  proverb  which  says:  "Legs  of  a  Caryota,^ 
legs  of  a  Pandanus;  women  hither,  women  thither."  The 
Pandanus  and  the  Caryota  are  alike  in  having  visible  aerial 
roots  diverging  downwards  to  the  ground  from  the  trunk; 
and  the  meaning  of  the  proverb  is  that  diverse  groups  are 
tied  together  by  affinal  bonds  due  to  past  marriages  of  pairs 
of  sisters.  The  genealogical  pattern  of  the  spreading  of  these 
links  is  represented  in  the  diagram.  Fig.  3. 

Two  sisters  of  clan  A  get  married.  One  of  them  marries 
into  clan  B  and  the  other  marries  into  clan  C.  Their  offspring 

^  This  identification  is  uncertain.  I  should  rather  have  said  "a  species 
of  palm  which  looks  like  a  Caryota"  but  this  phrase  could  hardly  be 
inserted  into  the  proverb. 


The  Sociology  of  Naven  95 

are  regarded  as  classificatory  siblings  although  the  children 
of  the  one  woman  are  of  clan  B  and  those  of  the  other  of 
clan  C.  In  this  filial  generation,  the  girls  will  again  marry 
into  other  clans,  e.g.  the  girl  of  clan  B  marries  a  man  of 
clan  E,  and  the  girl  of  clan  C  marries  a  man  of  clan  D.  The 
grandchildren  from  the  original  marriages  will  again  regard 
themselves  as  classificatory  siblings.  The  members  of  what 
is  now  the  middle  generation  will  become  classificatory 
"fathers",  "mothers",  "mother's  brothers"  and  "father's 
sisters  "  of  these  children :  so  that  owing  to  the  original  mar- 
riage of  two  sisters  of  clan  A  it  has  come  about  that  a  man  in 
clan  B  has  a  laua  in  clan  D  and  a  man  in  clan  C  has  a  laua 
in  clan  E. 

But  though  the  term  laua  nyanggu  may  be  applied  to  all 
these  classificatory  lauas  and  even  extended  to  apply  to  their 
fathers  and  other  members  of  their  clans,  a  man  will  take 
steps  to  mark  out  his  relationship  to  the  laua  groups  whose 
allegiance  he  especially  desires^  and  he  will  demonstrate  that 
he  really  is  a  wau  by  giving,  to  the  baby  laua,  a  name  of 
which  the  termination  means  "mask".  This  gift  of  a  name 
is  accompanied  by  the  gift  of  a  coconut  and  it  is  followed  in 
later  years  by  the  whole  naven  sequence.  The  zcau  will  make 
it  his  business  to  remind  the  laua  constantly  of  his  relation- 
ship ;  whenever  he  meets  the  boy  he  will  address  him  by  the 
name  which  he  gave  him,  he  will  exclaim  at  the  laua's  achieve- 
ments and  he  will  present  the  laua  with  meat.  The  laua  will 
give  return  presents  of  valuables  to  the  zvau,  and  will  come 
and  work  whenever  the  wau  is  gathering  all  his  lanoa  nampa 
and  laua  nyanggu  for  some  heavy  work  in  house-building  or 
clearing  a  garden. 

These  two  types  of  afiinal  relationship  were  even  more 
clearly  defined  by  one  informant,  in  Palimbai,  who  explained 
to  me  that  he  had  two  sorts  of  tawontu  (wife's  brother).  Of 
these  relatives,  one  kind  were  "the  men  who  had  received 
a  part  of  the  bride  price  he  gave  for  his  wife  ",  i.e.  the  members 
of  his  wife's  clan  of  her  generation.  The  other  series  were 

^  Unfortunately  I  did  not  collect  any  material  bearing  on  the  circum- 
stances likely  to  influence  a  wau  in  his  choice  of  one  classificatory  laua 
rather  than  another  for  this  recognition. 


96  The  Sociology  of  Naven 

"  tawontu  who  would  celebrate  naven  for  his  children  ".  Thus 
the  bride  price  serves  as  the  defining  factor  in  the  relation- 
ship with  one  group  of  affines,  and  the  system  of  naven  and 
names  given  by  the  maternal  clan  serves  as  the  defining  factor 
in  the  other  case,^  By  the  combination  of  these  two  ^  systems, 
the  community  is  so  knit  together  that  it  can  accomplish 
great  architectural  works  and  warfare. 

It  may  be  asked  why  the  culture  should  preserve  these  old 
afHnal  linkages  by  stressing  the  resulting  classificatory  wau- 
laua  relationships  rather  than  by  stressing  the  corresponding 
classificatory  lando-tawontu  relationships — and  this  in  spite 
(5f  the  fact  that  the  latter  would  seem,  to  a  European,  to  be 
the  simpler  and  more  direct.  The  answer  to  this  question  is, 
I  think,  to  be  found  in  the  fact  that  the  actual  system  fits 
with  the  remainder  of  the  culture,  e.g.  with  the  naming 
system  in  which  waus  give  names  to  their  lauas.  In  fact,  the 
answers  to  questions  of  this  type  are  to  be  sought  not  in 
sociological  terms  but  rather  in  terms  of  the  cultural  structure.^ 

The  reality  of  the  integrating  effect  of  naven  might  still 
be  doubtful  if  the  latmul  communities  were  limited  in  size 
by  some  factor  other  than  the  weakness  of  their  internal 
integration.  It  is  conceivable,  for  example,  that  a  community 
whose  size  was  limited  by  its  physical  environment  would 
never  reach  that  size  at  which  every  contributory  integrating 
factor  becomes  relevant.  But  actually,  in  latmul  culture,  it  is 

^  The  informant  who  defined  this  position  was,  I  think,  somewhat 
more  definite  in  his  statements  than  the  actual  facts  of  the  culture  justify. 
In  many  cases  a  child's  own  mother's  brother  will  give  it  a  name ;  and  it  is 
likely  that  there  is  some  variation  in  these  respects  from  one  village  to 
another. 

*  There  is  a  third  type  of  extension  of  the  affinal  system  which  is 
important  in  the  integration  of  ceremonial  and  architectural  undertakings 
in  the  ceremonial  house.  This  is  the  relationship  between  certain  pairs 
of  clans  which  regard  each  other  reciprocally  as  lanoa  nampa  or  laua 
nyanggu — both  terms  are  used — and  who  do  a  great  deal  of  work  for  each 
other.  This  relationship  is  reciprocal  between  clans  and  apparently  depends 
not  upon  any  particular  present  or  past  marriage,  but  upon  a  tradition 
that  the  women  of  the  one  clan  often  marry  the  men  of  the  other  and 
vice  versa.  The  relationship  finds  ritual  expression  on  various  occasions, 
especially  in  mortuary  feasts,  but  is  never,  I  think,  marked  by  the  naven 
system.  This  clan  relationship  is  practically  without  effect  outside  the 
ceremonial  house. 

'  Cf.  Epilogue,  p.  269,  where  another  type  of  answer  to  this  question 
is  suggested. 


The  Sociology  of  Naven  97 

clear  that  the  factor  which  limits  the  size  of  the  villages  is  the 
weakness  of  their  internal  cohesion.  The  larger  villages  are 
continually  on  the  point  of  fission,  and  the  fissions  which 
have  taken  place  in  the  past  are  always  ascribed  to  quarrels 
which  have  split  the  parent  community.  More  than  this, 
these  fissions  when  they  occur  invariably  follow  the  lines  of 
the  patrilineal  groups — a  clan,  a  phratry  or  a  moiety  splitting 
off  from  the  parent  community  and  thereby  rending  the 
system  of  affinal  linkages.  From  this  pattern  of  cleavage  it 
is  evident  that  the  patrilineal  linkages  are  stronger  than  the 
affinal,  the  latter  presenting  as  it  were  a  plane  of  weakness  of 
the  community.  Under  these  circumstances  the  importance 
of  any  factor  which  strengthens  the  affinal  links  becomes 
apparent  and  we  are  justified  in  saying  that  the  villages  could 
not  be  as  big  as  they  are  if  it  were  not  for  naven  ceremonies 
or  some  analogous  phenomenon. 

The  phrase,  "some  analogous  phenomenon",  might  well 
denote  a  mechanism  entirely  different  from  the  naven,  one 
even  which  did  not  rely  for  its  effect  upon  the  stressing  of 
the  affinal  links.  The  integration  of  society  might  for  example 
be  built  up  by  the  attachment  of  all  the  individuals  to  some 
chief,  their  allegiance  to  whom  would  then  be  strengthened 
by  ceremonial  of  some  quite  other  pattern.  Equally,  the 
same  sociological  function  might  be  achieved  by  any  of  the 
mechanisms  which  enforce  codified  law.^  Indeed  the 
sociological  position  of  naven  is  hardly  comprehensible  to  a 
European  until  he  has  some  picture  of  how  a  society  works 
without  either  officials  or  codes  of  law. 

The  latmul  are,  fundamentally,  a  people  without  law.   By 

^  Malinowski  has  stated  in  the  Introduction  to  Hogbin's  Law  and  Order 
in  Polynesia  that  an  analogue  of  our  own  "law"  may  be  found  in  all 
cultures,  and  this  is  no  doubt  true  if  we  take  a  sufficiently  abstract  view 
of  the  matter.  But  we  should  not  lose  sight  of  the  fact  that  the  only 
resemblance  between  the  European  legal  system  and  such  a  system  as 
that  of  the  latmul  lies  in  their  widest  sociological  functions.  We  may, 
for  example,  say  that  a  system  of  codified  law,  or  a  strong  chieftainship, 
performs  in  higher  cultures  many  of  the  functions  which  vengeance 
performs  among  the  latmul.  But  in  a  functional  study  we  cannot  afford 
to  ignore  the  mechanisms  involved — equating  jellyfish  with  fish  because 
they  both  swim.  To  the  writer  it  seems  that  there  is  a  very  profound 
contrast  between  peripherally  oriented  systems  such  as  that  of  the  latmul 
and  centripetal  systems  such  as  those  of  Western  Europe. 


98  The  Sociology  of  Naven 

this  I  do  not  mean  that  they  have  not  customs  and  sanctions, 
but  that  they  have  no  codified  law  and  no  estabHshed 
authority  which  might  impose  sanctions  ex  officio  or  in  the 
name  of  the  community  as  a  whole.  It  is  a  general  principle — 
though  one  not  stated  as  a  generalisation  by  the  natives  them- 
selves— that  no  considerable  sanction,  i.e.  no  fine,  damage  to 
property  or  physical  injury,  is  ever  imposed  upon  an  indi- 
vidual by  the  group  (clan,  moiety,  initiatory  group,  or  village) 
of  which  the  individual  is  himself  a  member;  neither  is  any 
considerable  sanction  ever  imposed  within  the  group  by  a 
superior  authority  representing  that  group  or  by  any  out- 
standing man  in  the  group.  In  general  such  internal  sanc- 
tions confine  themselves  to  disapproval,  insult  and  abuse,^ 

In  the  absence  of  such  sanctions,  internal  or  imposed  from 
above,  the  ordering  of  latmul  society  is  almost  entirely  de- 
pendent upon  what  we  may  call  external  or  lateral  sanctions. 
The  tendency  of  the  people  is  to  phrase  every  oflFence  as  an 
offence  against  somebody  and  to  leave  the  business  of  inflicting 
sanctions  to  the  offended  people. 

In  its  simplest  forms  the  system  works  thus :  A  offends  B. 
B  avenges  himself  with  weapons  or  sorcery.  Alternatively  B 
may  produce  such  a  show  of  anger  and  power  or  such  a  shower 
of  abuse  that  A,  either  frightened  or  ashamed,  will  pay  some 
sort  of  compensation  in  shell  valuables,  pigs  or  areca  nut. 
B  will  then  formally  accept  this  compensation,  either  stepping 
on  the  pig,  standing  while  A  puts  the  bunch  of  areca  nuts 
on  his  shoulder,  or  permitting  A  to  cut  his  hair,  oil  his  face 
and  otherwise  decorate  him.  In  some  cases,  B  will  finally 
make  a  presentation  of  some  sort  to  A. 

This  phrasing  of  the  matter  in  individualistic  terms  is,  as 

^  The  only  exceptions  to  this  rule  are :  {a)  Cases  in  which  an  outstanding 
member  of  the  group  will  recommend  the  offender  to  compensate  an 
offended  person  outside  the  group.  Such  recommendations  may  be  angry 
and  vociferous  but  they  remain  mere  recommendations  and  are  in  no 
sense  a  decision  of  the  dispute  by  constituted  authority.  The  outstanding 
man  speaks  as  one  who  is  concerned  in  the  dispute  and  who  may  be 
seriously  involved  if  the  quarrel  grows.  He  is  not  an  impartial  outsider. 
(6)  Cases  inside  the  family  group.  Here  the  husband  has  authority  over 
his  wives ;  if  two  wives  are  jealous  he  may  beat  them  both.  Father  has 
authority  over  son  and  elder  brother  over  younger  brother.  But  this 
pattern  is  confined  to  the  family  itself  and  is  not  repeated  in  the  larger 
groups  or  in  the  society  as  a  whole. 


The  Sociology  of  Naven  99 

we  have  already  seen  in  the  chapter  on  "Sorcery  and  Ven- 
geance", to  some  extent  obscured  by  the  phenomena  of 
identification  whereby,  when  A  offends  B,  the  resulting 
quarrel  is  taken  up  by  their  relatives,  fellow  clan  members, 
or  members  of  the  same  initiatory  groups.  But  always  the 
offender  expects  his  own  people  to  be  active  on  his  side  and 
the  quarrel  remains  one  between  two  peripheral  groups,  never 
taking  the  centripetal  form  which  we  might  phrase  as  "  Rex 
or  State  versus  So-and-so".  Such  a  form  is  indeed  pre- 
cluded by  the  considerable  feeling  which  exists  in  the  latmui 
community  against  any  participation  of  outside  persons  in 
quarrels  which  do  not  concern  them.^ 

The  whole  mechanism  of  latmui  sanctions  differs  so  pro- 
foundly from  our  own,  that  I  must  illustrate  it  with  examples : 

1 .  My  informant,  Tshava,  of  Kankanamun,  was  a  man  of 
about  thirty-five.  The  following  incident  occurred  when  he 
was  a  small  boy :  Another  boy,  Tshaguli-mbuangga,  had  been 
stealing  Tshava's  things  from  his  house.  Tshava  complained 
of  this  to  his  own  father.  The  father  watched  for  Tshaguli- 
mbuangga  and  when  he  saw  him  enter  the  house  he  gave  a 
signal  to  Tshava.  Tshava  then  killed  Tshaguli-mbuangga 
with  a  spear.   His  father  helped  him. 

In  Tshava's  account  of  the  matter  there  is  no  mention  of 
any  annoyance  on  the  part  of  the  murdered  boy's  relatives, 
but  this  no  doubt  occurred.  In  any  case  Tshava  presented 
valuables  to  them.  He  states  that  he  paid :  One  Turbo  shell, 
one  mother-of-pearl  crescent,  one  tortoise-shell  arm  band, 
and  one  necklace  of  small  Conus  shells. 

The  relatives  of  the  boy  whom  he  killed  presented  a 
tambointsha,  i.e.  a  tassel,  to  Tshava.  This  he  will  later  exhibit 
on  his  lime  stick  (cf.  Plate  XXIVa)  as  a  symbol  of  successful 
homicide.  But  as  yet  his  initiatory  grade  have  not  been 
promoted  to  the  wearing  of  this  type  of  ornament. 

2.  While  I  was  in  the  village  of  Komindimbit  watching 
initiation  ceremonies,  a  woman,  a  visitor  from  another  village, 
was  detected  spying  upon  the  flutes  from  a  Malay  Apple  tree 
up  which  she  had  climbed.  I  heard  a  lot  of  shouting  in  the 
ceremonial  house  and  dashed  to  the  scene.  I  found  a  number 

^  Cf.  Case  9,  p.  69. 


100  The  Sociology  o/Naven 

of  men  jumping  about  with  sticks  in  their  hands  while  others 
remained  quietly  sitting  on  the  platforms.  The  active  angry 
group  were  the  members  of  the  moiety  B,  especially  Bx  4 
and  By  4  (cf.  Diagram,  p.  245).  These  were  the  groups  which 
had  just  paid  for  the  privilege  of  seeing  the  flutes  and  there- 
fore the  offence  was  regarded  as  specifically  against  them, 
not  as  against  the  whole  group  of  men  of  that  ceremonial 
house.  One  man  of  moiety  A  did,  it  is  true,  give  a  many- 
pronged  spear  to  a  member  of  moiety  B,  saying,  "Go  on, 
spear  her ! "  But  this  was  the  only  part  they  took  in  the  matter. 
A  party  of  the  angry  young  men  dashed  off  with  sticks  in 
their  hands  to  search  the  village  for  her.  But  they  returned 
without  beating  her — I  think  they  could  not  find  her.  The 
remainder  of  the  group  contented  themselves  with  dancing 
and  jumping  about  in  the  ceremonial  house  with  sticks  and 
spears. 

Finally  the  members  of  moiety  A  quieted  them  and  per- 
suaded them  to  accept  compensation.  The  two  leaders  or 
nambu  wail  (literally  "heads  of  crocodiles" — an  initiatory 
grade  is  called  a  "crocodile")  oi  Ax  4  and  By  4  accordingly 
stood  while  bunches  of  areca  nuts  with  valuables  attached 
were  placed  upon  their  shoulders.  The  valuables  were:  two 
mother-of-pearl  crescents  and  two  Turbo  shells.  These  would 
later  be  used  in  buying  a  pig  for  some  feast. 

3.  While  I  was  in  Kankanamun  a  scandal  occurred  in  the 
mbwole  or  junior  ceremonial  house.  A  man  named  Wi-ndjuat- 
mali  saw,  in  the  middle  of  the  night,  a  couple  copulating 
there.  He  identified  them  as  T.  and  T.'s  wife.  It  was  none 
of  Wi-ndjuat-mali's  business  because  he  belongs  to  the  group 
B  4  and  therefore  is  a  member  of  the  big  ceremonial  house 
and  has  nothing  to  do  with  the  mbwole  which  belongs  especially 
to  the  groups  Ax  5  and  Ay  5.  But  there  is  a  great  deal  of 
rivalry  between  the  moieties  A  and  B  and  it  is  possible  that 
Wi-ndjuat-mali,  as  a  member  of  B,  took  a  certain  malicious 
pleasure  in  telling  the  A  group  of  the  misfortune  which  had 
overtaken  their  mbwole.  In  any  case  Wi-ndjaut-mali  told  a 
member  oi  A  $  what  he  had  seen  and  the  latter  immediately 
told  the  rest  of  ^  5.  T.,  the  culprit,  was  himself  a  member 
of  this  group  and  thus  it  came  about  that  a  row  was  pre- 
cipitated inside  the  group  A  5. 


The  Sociology  of  Naven  loi 

^T,  was  summoned  by  means  of  the  slit-gong  in  the  mbwole 
but  he  did  not  come.  Nothing  was  done  till  the  evening. 
Then  the  gong  was  beaten  again  and  T.  came.  Then  they 
debated  against  him:  "Why  did  you  copulate  in  our  mbwole} 
You  have  made  us  'cold'.  One  and  all  we  are  utterly  bad; 
and  never  again  shall  we  be  all  right.  Another  time  if  you 
stay  in  your  house  (not  answering  the  gong)  we  shall  go  up 
into  it  and  beat  you  there.  We  shall  put  spells  upon  your 
wife  so  that  she  runs  away  to  another  man." 

Then  T.  said,  "Make  Wi-ndjuat-mali  come  quickly.  They 
are  joking;  they  are  lying  against  me."  But  Wi-ndjuat-mali 
said,  "  It  is  no  lie,  it  is  true." 

Then  they  said,  "Your  wife — we  will  kill  a  pig  (and  pay 
for  it)  with  her  property.  We  are  children  with  no  property. 
When  the  elder  grade  call  for  a  pig,  we  will  kill  it  with  his 
property. 

"She  is  not  a  woman  who  copulates  with  a  small  vulva. 
Her  secretions  have  spilt  in  our  mbwole.  And  she  is  a  preg- 
nant woman  too.  Yes.  The  semen  that  you  have  ejaculated 
will  cover  up  the  child.  They  will  have  to  scrape  the  child 
when  it  is  bom."^  And  so  on. 

A  deaf  and  dumb  man  also  took  part  in  this  debate  and 
expressed  himself  in  noises  and  gestures  with  great  force ;  but 
his  remarks  were  untranslatable. 

Of  the  various  speakers  against  T.  in  this  debate,  all  were 
members  of  moiety  A  with  the  exception  noted  in  the  text 
of  Wi-ndjuat-mali  (B  4)  who  spoke  only  as  a  witness.  Most  of 
the  speakers  were  members  of  ^  5,  but  one  or  two  men  who 
were  half-promoted  from  A  5  into  A  3  also  spoke.  The 
debate  went  on  far  into  the  night. 

A  few  days  later  I  asked  some  questions  about  the  incident. 
To  my  question,  "Why  did  T.  and  his  wife  use  the  mbwole 
when  they  had  a  house  of  their  own?"  I  received  the  reply, 
"He  and  his  wife  had  a  long  love  affair  before  they  were 
married.  They  used  then  to  use  all  sorts  of  places  and  so  they 
still  go  on  doing  the  same  thing." 

^  The  following  is  a  translation  of  a  dictated  account  of  this  debate. 
Explanatory  phrases  not  in  the  original  have  been  inserted  in  brackets. 

'  Whenever  this  happens  it  is  suggested  that  the  parents  failed  to  keep 
the  taboo  on  copulation  during  pregnancy. 


102  The  Sociology  of  Naven 

To  my  question,  "Are  they  going  to  kill  one  of  T.'s  pigs 
or  take  any  of  his  property? "  the  reply  was  "  No.  Of  course 
not.   He  is  a  member  of  their  own  initiatory  group." 

This  last  statement  was  the  crux  of  the  matter  and  the  lack 
of  any  punitive  mechanism  was  due  simply  to  this  fact.  But 
such  cases  in  which  no  sanction  can  be  applied  are  com- 
paratively rare.  The  situation  was  exceptional  owing  to  the 
whole  scandal  being  connected  with  the  mbzvole  where  only 
one  of  the  two  moieties  A  and  B  is  represented  and  where  also 
the  moieties  x  and  y  are  not  sharply  pitted  against  each  other. 
Had  a  corresponding  outrage  been  committed  in  the  big 
ceremonial  house  the  resulting  row  would  probably  have  been 
between  the  moieties  A  and  B,  the  various  grades  of  each 
moiety  taking  sides  against  the  opposite  moiety.  Other  types 
of  segmentation  might,  however,  have  occurred  in  the  big 
ceremonial  house.  The  building  is  subdivided  among  the 
various  clans  and  the  moieties  x  and  y.  Thus  the  oifence,  if 
it  occurred  in  a  particular  part  of  the  ceremonial  house, 
might  have  been  phrased  as  an  offence  against  a  particular 
clan,  or  as  an  offence  of  x  against  y. 

4.  While  I  was  in  Palimbai  a  quarrel  broke  out  in  a  house 
which  was  shared  by  two  classificatory  brothers,  Koulavw^an 
and  Membi-awan  (cf.  Genealogy).  Both  of  these  brothers 
were  married,  but  the  wife  of  Membi-awan  had  recently  borne 
a  child  which  she  was  suckling.  Membi-awan  was  therefore 
unable  to  sleep  with  her.  Koulavwan  went  away  for  one 
night  to  trade  with  a  bush  village  leaving  his  wife,  Kaindshi- 
mboli-agwi,  behind  in  Palimbai.  On  his  return,  the  following 
afternoon,  he  was  told  by  his  mother  Kapma-tshat-tagwa  and 
the  other  women  of  the  house  that  Membi-awan  had  gone 
in  the  night  to  Kaindshi-mboli-agwi's  mosquito  bag. 

It  appears  that  what  actually  happened  in  the  night  was 
as  follows:  Membi-awan  went  to  Kaindshi-mboli-agwi's  bag, 
opened  the  end  of  it  and  put  his  hand  in  and  there  felt  her 
head.  She  woke  up  and  shouted  and  Kapma-tshat-tagwa  came 
out  of  her  bag  to  see  what  was  the  matter.  Membi-awan  then 
hid  among  the  pots  in  the  house ;  but  he  was  identified. 

This  led  immediately  to  three  more  or  less  separable 
quarrels : 


The  Sociology  of  Naven  103 

{a)  Between  the  two  classificatory  brothers  Koulavwan  and 
Membi-awan.  Koulavwan  took  his  complaint  to  the  village 
"Tultul",  a  native  appointed  by  the  Government  to  act  as 
an  interpreter  to  the  District  Officer.  This  official  heard  what 
both  parties  had  to  say  but  did  nothing  about  it.  Membi-awan 


^D 


1-0+ 


-Of 
II 

^D 

CM- 

II 
04' 


Of 


r-  '^^KWOSHIMBA 


\) 


KANDAM-NOWE 
(my  informant) 


—  *^  WOLI-YIMBANG 

II     

CH- 


^WOMPUR-NDEMI 

II 

0+  lAVTAGWA 

.  o+Kj\Jndshi-mboli-agwi 


•Of  KAPMA-TSHAT-TAGWA 


r-"0 


KILI-MALI 


'^  KOULAVWAN 


Of-  MWAINDSHl-NYAUT 


Of  NAUA-RAGWA 

Of 
4 


0+ 

II  

U  MEMBI-AVWAN 


viS 


0+ 


claimed,  in  an  attempt  at  self-justification,  that  he  had  that 
day  seen  Kaindshi-mboli-agwi  give  a  betel  leaf  to  her  sister's 
husband,  Wompur-ndemi,  and  that  his  visit  to  her  mosquito 
bag  was  in  order  to  verify  that  his  brother's  wife  was  being 
faithful.  This  excuse  was  brushed  aside.  She  said,  "You 
wanted  to  copulate  with  me.   Now  you  are  lying." 


104  -^^  Sociology  of  Naven 

Koulavwan  insulted  Membi-awan,  saying,  "Yes,  when  we 
went  to  work  for  the  white  man,  you  were  only  a  cook^ 
but  I  worked.'* 

(b)  Between  Koulavwan  and  his  wife.  He  beat  her  "be- 
cause she  had  not  herself  come  and  told  him,  but  had  waited 
till  Kapma-tshat-tagwa  told".  She  tried  to  placate  him  by 
saying,  "My  husband  never  copulates  with  other  people; 
I  am  going  to  be  like  that.  My  husband  is  straight  and  I  am 
the  same."  Further,  she  pointed  out  that  she  had  expected 
something  of  the  sort  to  happen  and  had  asked  the  litde  girl, 
Mwaindshi-nyaut,  to  come  and  sleep  with  her  in  her  bag  for 
protection. 

But  my  informant,  her  classificatory  brother,  said  to  Kou- 
lavwan, "  Go  on  and  beat  my  sister ;  she  ought  to  have  hurried 
up  and  told  you."   So  Koulavwan  beat  her. 

(c)  This  quarrel  between  Koulavwan  and  his  wife  had 
reverberations  in  another  family :  Kili-mali,  sympathising  with 
Koulavwan,  started  to  beat  his  own  wives  and  children, 
saying,  "They  were  all  asleep.  They  ought  to  have  heard 
what  was  happening." 

These  three  quarrels  apparently  died  down  before  night 
and  all  the  persons  concerned  went  to  bed  as  usual. 

But,  in  the  middle  of  the  night,  Koulavwan  got  up,  angry, 
out  of  his  mosquito  bag  and  went  and  jumped  on  Membi- 
awan's  bag  in  which  the  latter  was  sleeping.  Then  the  two 
men  fought. 

After  that,  still  dissatisfied,  Koulavwan  started  to  beat 
Kaindshi-mboli-agwi  again.  Both  husband  and  wife  fell  into 
the  water  under  the  house — it  was  then  the  flooded  season — 
and  she  screamed  for  Kandem-nowe  who  was  sleeping  in  the 
next  house  but  one.  He  dashed  to  the  scene  to  defend  her — 
"We  are  both  of  the  same  nggwotl-nggu."^ 

ThenWoli-yimbang  and  Kwoshimba  came  on  the  scene  and 

^  I.e.  a  cook  for  the  labour  line.  The  office  of  cook  in  a  European's 
house  is  superior,  but  the  cook  for  the  labourers  is  despised. 

*  This  term  is  applied  to  a  close  patrilineal  group  within  the  clan.  The 
word  is,  I  think,  a  plural  of  the  word  nggwail  (father's  father  or  son's  son), 
an  exception  to  the  usual  rule  according  to  which  the  latmul  language 
lacks  plural  suffixes.  The  only  other  similar  words  that  I  know  of  are: 
nyan-nggu  (children)  and  mbwambo-nggu  (mother's  fathers). 


The  Sociology  of  Naven  105 

began  to  take  Kaindshi-mboli-agwi's  part.  They  scolded  Kou- 
lavwan — "  Kaindshi-mboli-agwi  is  your  mother.  You  ought 
to  hit  her  more  gently."  (This  terminology  is  based  on  the 
fact  that  Kaindshi-mboli-agwi  is  Woli-yimbang's  daughter  and 
Woli-yimbang  is  a  classificatory  brother  of  Kapma-tshat- 
tagwa.  Therefore  Kaindshi-mboli-agwi  is  Koulavwan's  zoau's 
daughter,  a  relative  who  is  called  nyame^  mother.) 

Woli-yimbang  took  a  paddle,  and  Kwoshimba  took  the 
stalk  of  a  sago  leaf.  With  these  weapons  they  made  a  show  of 
beating  Koulavwan. 

Then  Kili-mali,  Koulavwan's  father,  came  up  and  was  angry 
with  Woli-yimbang  and  Kwoshimba.  He  said  they  had  no 
right  to  interfere.  "  Kaindshi-mboli-agwi  is  no  longer  your 
daughter.    She  belongs  to  us." 

In  reply  the  two  fathers  of  Kaindshi-mboli-agwi  at  once 
started  to  complain  about  the  quality  of  the  valuables  that 
had  been  paid  to  their  clan  as  a  bride  price  for  Kaindshi- 
mboli-agwi  ;  and  Kili-mali  replied  with  comments  upon  the 
nggwat  keranda  (presentation  of  valuables  from  the  bride's 
people  to  those  of  the  husband).  This  led  to  general  re- 
criminations on  the  subject  of  their  mutual  indebtedness. 
Kwoshimba  pointed  out  that  he  was  always  helping  Kili-mali 
with  his  ngglambi  (i.e.  paying  out  valuables  so  that  Kili-mali 
might  make  sacrifices  to  the  zvagan).  "In  fact",  said  Kwo- 
shimba, "I  am  your  elder  brother." 

Kili-mali  denied  all  this  indebtedness  and  said  that  Kwo- 
shimba had  always  been  a  poor  man  till  Koulavwan  married 
Kaindshi-mboli-agwi  and  gave  him  some  mother-of-pearl 
crescents. 

Kwoshimba  then  referred  to  an  incident  in  the  distant 
past,  when  he  had  half-killed  a  man  and  permitted  Kili-mali 
to  complete  the  homicide  and  claim  the  achievement:  "I 
really  killed  that  man,  but  you  wore  the  homicidal  paint  for 
him.   How  many  have  you  killed  for  yourself.'' " 

Finally  Koulavwan  intervened  between  Kili-mali  and  Kwo- 
shimba, saying,  "  Our  fathers  must  not  begin  fighting."  The 
disputants  then  ceased,  and  Kwoshimba  and  Woli-yimbang 
said,  "True,  there  is  our  blood  in  Koulavwan"  (i.e.  "Kou- 
lavwan is  related  to  us  through  his  mother").    Then  they 


io6  The  Sociology  of  Nsven 

exchanged  areca  nuts  with  Kili-mali  as  a  sign  of  reconcilia- 
tion ;  and  all  went  back  to  bed. 

These  four  examples  of  the  latmul  handling  of  offences, 
together  with  those  which  have  been  given  in  the  chapter  on 
"  Sorcery  and  Vengeance",  will  suffice  to  show  the  principles 
and  diffuse  nature  of  the  system.  In  the  present  context,  that 
of  enquiry  into  the  sociological  functions  of  affinal  links,  the 
conclusions  which  I  draw  from  this  material  may  be  phrased 
as  follows : 

In  every  society,  divergences  from  the  cultural  norm  are 
liable  to  threaten  the  integration  of  that  society,  and  this  is 
clearly  evident  among  the  latmul.  But  under  the  latmul 
system  the  type  of  disintegration  which  is  threatened  is  some- 
what different  from  that  which  threatens  our  own  societies. 

In  our  own  case,  the  threat  is  one  of  increasing  muddle 
and  confusion  in  our  cultural  norms,  a  rot  which  might 
spread  through  the  community,  producing  first  a  lack  of 
orientation  among  the  individuals  and  finally  the  collapse  of 
the  society.  Occasionally,  in  cases  where  the  divergence  has 
some  positive  quality  bringing  with  it  a  new  orientation,  the 
divergent  individuals  unite  together  to  form  a  group  within 
the  major  community,  a  group  with  different  cultural  norms 
from  the  rest  of  the  population.  When  this  occurs,  society 
can  often  accommodate  the  resulting  group  provided  only 
that  it  is  not  too  active  in  its  opposition  to  the  centres  of  the 
centripetal  system.  Alternatively,  the  divergent  group  may 
either  succeed  in  altering  the  cultural  system  of  the  society 
or  it  may  go  off  to  some  other  part  of  the  world  and  there 
found  a  culture  different  from  that  of  the  parent  group. 

But  among  the  latmul  the  position  is  very  different.  Here 
there  can  be  no  threat  of  general  confusion  resulting  from  a 
blurring  of  cultural  norms,  since  the  chronic  state  of  the 
culture  is  one  in  which  the  norms  are  weakly  defined.^ 
Neither  can  there  be  a  conflict  between  an  organising  centre 
and  divergent  groups,  because  there  is  no  organising  centre 
in  the  purely  peripheral  system  of  the  latmul.  Rather  the 
threat  is  one  of  fission  of  the  community. 

^  The  extremely  drastic  circumstances  of  modem  culture  contact  with 
the  white  man  have,  however,  had  an  effect  somewhat  of  this  nature. 
Cf.  the  story  of  Tshimbat,  p.  i68. 


The  Sociology  of  Naven  107 

We  have  seen  how  every  quarrel  leads  to  the  integration 
of  loosely  defined  rival  groups  around  the  protagonists.  It  is 
this  process  which  continually  threatens  the  integration  of 
the  latmul  village.  With  more  profound  rivalry  the  groups 
may  grow  until,  in  extreme  cases,  the  fission  actually  takes 
place.  One  of  the  contending  groups  then  leaves  the  village 
and  founds  a  colony  elsewhere,  but  with  the  same  set  of 
norms  as  the  parent  community. 

A  further  diff^erence  between  the  two  systems  is  implicit 
in  what  I  have  already  said :  in  our  own  community  the  sub- 
groups which  split  from  the  parent  are  usually,  perhaps 
always,  formed  of  individuals  whose  divergent  behaviour  is 
based  upon  some  form  of  doctrine.  The  danger  to  the  status 
quo  in  our  communities  is  the  man  who  discards  the  cultural 
norms  and  who,  either  articulately  or  by  example,  "teacheth 
other  men  so  ".  Among  the  latmul,  who,  it  seems,  normally 
regard  the  rules  as  things  to  break  if  you  are  strong  enough, 
this  situation  does  not  arise.  Their  fissions  spring  not  from 
conflicting  doctrines  but  from  rivalry  between  individuals  or 
groups. 

Such  then  is  the  danger  which  continually  threatens  the 
latmul  community;  and  when  this  sociological  position  is 
stated  the  relevance  of  the  stressing  of  the  affinal  links  to  the 
integration  of  the  community  is  at  once  evident.  These  links 
form  a  network  which  runs  across  the  patrilineal  systems  of 
clans,  moieties,  and  initiatory  groups  and  thereby  ties  the 
conflicting  groups  together.  More  than  this,  the  patterning 
and  ubiquity  of  the  affinal  relationships  are  such  as  to  ensure 
that  whenever  a  quarrel  reaches  serious  dimensions  there 
shall  be  some  individuals  marked  out  to  act  as  peacemakers, 
to  intervene  between  the  disputants  as  Koulavwan  in  the 
example  given  above  intervened  between  his  father  and  his 
wan. 


CHAPTER  VIII 

Problems  and  Methods  of  Approach 

Problems.  In  our  endeavour  to  relate  the  details  of  the  naven 
to  their  setting,  we  examined  first  the  ceremonies  themselves, 
then  the  basic  formulations  upon  which  the  wau  relation- 
ship is  built,  and  lastly  the  function  which  the  ceremonies 
perform  in  integrating  the  Society.  The  structural  and  socio- 
logical analysis  has  revealed  two  important  facts:  (i)  the 
cultural  structure  provides  a  setting  with  which  the  cere- 
monies are  consistent;  and  (2)  the  ceremonies  perform  an 
important  function  in  stressing  certain  classificatory  relation- 
ships and  thereby  contribute  to  the  integration  of  latmul 
society;  crudely  stated — if  there  were  no  naven  ceremonies 
and  no  sociological  analogues  of  naven,  the  villages  would 
not  be  so  big.  We  have  then  two  sorts  of  factors,  structural 
and  sociological,  which  are  certainly  contributory  to  the  naven 
complex. 

But  in  the  study  of  equilibria — and  all  cultural  and  social 
anthropology  is  a  study  of  equilibria — we  can  never  be  sure 
that  all  the  relevant  factors  have  been  mentioned.  It  is 
exceedingly  dangerous  to  point  to  certain  factors  and  to  say 
that  they  constitute  the  whole  cause  of  the  effect  which  we 
want  to  explain.  In  the  present  context  it  would  have  been 
dangerous  to  point  to  the  structural  facts  upon  which  the 
naven  complex  is  built  and  to  say  that  the  sociological  aspects 
of  the  matter  were  irrelevant,  and  equally  dangerous,  in- 
tolerantly preferring  the  sociological  view,  to  deny  the  im- 
portance of  the  structure.  In  such  cases  it  is  safer  to  say  that 
the  known  factors  contribute  to  the  effect  in  question,  and  to 
go  on  with  the  search  for  other  types  of  factors  which  may 
also  contribute  their  quota  to  the  equilibrium.  It  may  often 
be  difficult  to  decide  whether  all  the  details  of  an  effect  can 
be  ascribed  to  the  known  causes,  but  in  the  present  instance 
there  are  many  points  which  are  not  adequately  explained 
by  either  the  structural  or  the  sociological  factors. 

In  the  first  place,  there  are  still  features  inside  the  naven 


Problems  and  Methods  of  Approach  109 

plexus  itself  for  which  we  have  not  accounted.  We  still  do 
not  know  why  the  antics  of  the  wau  should  be  so  exaggerated, 
nor  why  they  should  be  comic.  Why  in  stressing  his  identifica- 
tion with  the  boy's  mother  should  he  dress  himself  as  a  filthy 
and  disgusting  old  hag.''  Why  does  he  put  on  the  very  worst 
of  feminine  garments  while  the  father's  sister  puts  on  the 
very  best  male  ornaments  that  she  can  find? 

Secondly,  there  is  the  group  of  questions  raised  the  mo- 
ment we  push  our  enquiry  farther  along  the  branches  in  this 
network  of  cause  and  effect.  Our  theory  of  reasons  for  the 
development  of  naven  has  been  built  up  inside  a  closed 
system  in  which  the  size  of  the  village,  for  example,  was 
assumed  ex  hypothesi.  But  a  stable  culture  is  itself  a  complete 
functional  system  and  therefore  we  cannot  assume  certain 
facts  as  fixed  in  this  way,  but  must  still  try  to  relate  them  back 
to  the  other  features  of  the  culture  until  we  have  a  com- 
pletely circular  or  reticulate  exposition  of  causes  and  effects. 
We  must  therefore  ask:  Why  should  the  villages  be  so 
large  ? 

This  question  cannot  be  answered  out  of  hand,  in  the 
present  instance,  by  a  reference  to  head-hunting  warfare  as 
the  factor  which  puts  a  premium  on  large  size  of  village  and 
close  integration  of  the  community.  So  far  as  the  latmul 
culture  itself  is  concerned,  I  might  have  been  tempted  to 
regard  warfare  as  the  complete  answer  to  the  question,  but 
from  what  Dr  Fortune  tells  me  of  the  Mundugumor  people 
on  the  Yuat  River,  a  tributary  of  the  Sepik,  warfare  can  be 
at  best  only  a  contributory  factor,  since  the  Mundugumor 
people  combine  the  particular  traits  which  we  are  tempted 
to  regard  as  functionally  incompatible.  They  live  in  small 
conmiunities  which  are  further  subdivided  into  small  family 
hamlets ;  the  integration  of  the  community  is  at  so  low  a  level 
that  individuals — even  own  brothers — are  intensely  jealous 
and  hostile  among  themselves ;  lastly,  they  are  keen  and  suc- 
cessful head-hunters  and  cannibals. 

But  more  significant  than  either  of  the  above  types  of 
question  are  the  problems  of  individual  motivation.  Even 
if  we  were  studying  a  community  in  which  "civic  sense" 
was  very  highly  developed  in  the  individuals,  it  is  very 


no  Problems  and  Methods  of  Approach 

doubtful  whether  we  should  be  justified  in  regarding  the 
sociological  functions  of  their  behaviour  as  indications  of  the 
underlying  motives.  In  any  case  the  latmul  are  but  little 
given  to  expressing  the  reasons  for  their  behaviour  in  socio- 
logical terms. 

Similarly,  if  we  were  studying  a  people  who  set  great  value 
upon  the  logical  consistency  of  their  behaviour  we  should 
hesitate  before  accepting  their  logical  reasons  for  behaviour 
as  honest  statements  of  motive.  We  may  well  look  askance  at 
the  father  who  tells  us  that  he  is  beating  his  child  either  "  for 
the  good  of  the  State"  or  "because  it  is  the  logical  thing  to 
do". 

But  though  neither  structural  nor  sociological  phrasings 
can  be  accepted  as  statements  of  motive,  one  fact  has  been 
mentioned  in  passing  which  we  might  accept  as  a  sufficient 
statement  of  the  zvau's  motives  for  performance  of  naven. 
We  have  seen  that  as  a  result  of  naven  the  wau  gains  the 
allegiance  of  his  laua  and  we  might  dismiss  the  problem  as 
solved  in  those  terms. 

Such  a  solution  assumes,  however,  that  the  wau  wants 
allegiance,  and  to  me  this  assumption  is  unsatisfactory.  Does 
he  want  allegiance  because  all  male  human  beings  are  so 
born  that  they  must  later  crave  for  the  allegiance  of  their 
fellows?  Or  has  the  wau  been  so  moulded  by  the  culture 
in  which  he  has  lived  that  he  wants  allegiance? 

If  we  discard  the  genetical  view  of  the  matter  and  suppose 
that  the  desire  for  allegiance  is  a  conditioned  appetite  im- 
planted in  individuals  by  their  culture,  it  then  follows  that 
this  desire  may  take  very  various  forms  in  different  cultures. 
The  possession  of  many  loyal  supporters  may  be  associated 
with  the  most  various  forms  of  gratification.  For  example, 
an  individual  may  want  the  allegiance  of  his  fellows  because 
he  is  conditioned  to  feel  an  agreeable  emotional  warmth 
when  he  and  his  friends  are  all  jolly  together.  Alternatively, 
he  may  be  so  conditioned  that,  lacking  loyal  friends,  he  feels 
anxious  and  afraid  of  a  spear  in  his  back,  while  their  presence 
gives  him  confidence.  Or  again,  the  possession  of  loyal 
friends  may  be  a  conditional  stimulus  for  the  reactions  asso- 
ciated with  the  gratification  of  pride.  The  crude  assumption 


Problems  and  Methods  of  Approach  1 1 1 

that  all  men  desire  the  allegiance  of  their  fellows  ignores  all 
these  various  possibilities  and  so  tells  us  virtually  nothing 
about  the  wau's  motives. 

At  this  point  we  may  note  that  all  the  questions  which  have 
been  raised  in  this  chapter  are  closely  inter-related.  We  have 
asked:  (i)  Why  is  the  wau  a  buffoon?  (2)  Why  are  the 
villages  large?  (3)  What  are  the  wau's,  motives?  All  the^e 
questions  might  be  answered  in  terms  of  emotional  satis- 
faction given  to  the  individuals  by  the  various  phenomena, 
buffoonery,  large  villages  and  naven. 

But  we  have  noted  in  an  earlier  chapter  (p.  32)  that  the 
study  of  affective  functions  is  beset  with  difficulty  and  entails 
a  preliminary  study  of  the  ethos  of  the  culture ;  it  will  there- 
fore be  useful  at  this  point  to  examine  the  ethological  method 
in  rather  more  detail  and  to  consider  the  relationship  between 
ethology  and  the  concepts  of  philosophical  history  from  which 
it  is  derived. 

Zeitgeist  and  Configuration.  The  historical  approach  to 
the  cultures  of  primitive  peoples  has  been  frequently  abused 
by  students  who  practise  the  methods  of  functional  anthro- 
pology. It  has  been  said  that  the  historians  are  solely  con- 
cerned with  the  search  for  origins  and  with  the  construction 
of  speculative  narrative.  This  abuse  is  only  justified  if  the 
writings  of  the  historians  emphasise  this  aspect  of  their  work 
at  the  expense  of  more  scientific  aspects.  History,  in  so  far 
as  it  is  a  science,  is  concerned  not  with  narratives  and  origins 
but  with  generalisations  from  narrative,  generalisations  based 
upon  the  comparative  study  of  the  processes  of  cultural  and 
social  change.  The  major  achievement  of,  for  example,  the 
heliolithic  historians  is,  not  their  theory  that  almost  all  the 
cultures  of  the  world  have  been  derived  from  those  of  Egypt 
and  Sumeria,  but  the  picture  which  they  have  given  us  of 
the  processes  of  change  and  degradation  in  culture.  It  is 
indeed  high  time  that  some  student  set  to  work  upon  the 
classification  of  these  processes. 

In  the  description  of  diachronic  process,  historians  have 
used  many  of  the  same  functional  and  economic  concepts 
as  are  used  in  synchronic  studies ;  but  they  have  also  elabo- 
rated one  concept  which  has  only  very  recently  been  adopted 


112  Problems  and  Methods  of  Approach 

into  the  vocabulary  of  synchronic  anthropology.  This  is  the 
concept  of  Zeitgeist^  the  spirit  of  the  times,  a  concept  wjiich 
owes  its  origin  to  the  Dilthey-Spengler  school  of  philo- 
sophical history. 

The  suggestion  of  this  school  is  that  the  occurrence  of 
cultural  changes  is  in  part  controlled  by  some  abstract 
property  of  the  culture,  which  may  vary  from  period  to 
period  so  that  at  one  time  a  given  change  is  appropriate  and 
occurs  easily  though  a  hundred  years  earlier  the  same  innova- 
tion may  have  been  rejected  by  the  culture  because  it  was  in 
some  way  inappropriate. 

Dr  Benedict^  has  developed  a  related  concept,  that  of  the 
"configuration"  of  culture,  and  has  done  some  very  inter- 
esting and  important  work  on  primitive  cultures.  She  has 
shown,  for  instance,  that  the  refusal  of  the  Zuni  to  adopt 
either  peyote  or  alcoholic  drinks  was  conditioned  by  the 
Apollonian  configuration  of  their  culture,  while  the  neigh- 
bouring peoples  with  Dionysian  cultures  adopted  both  these 
stimulants  with  enthusiasm. 

In  European  history  the  same  type  of  concept  is  invoked 
in  order  to  explain  such  curious  facts  as  that  Leonardo's 
mechanical  inventions  passed  unnoticed  or  were  laughed  at 
during  his  life-time,  and  their  importance  was  only  gradually 
recognised  during  the  following  three  centuries;  or  that  the 
theory  of  evolution,  though  stated  many  times,  did  not  obtain 
general  acceptance  until  the  industrial  revolution  had  made 
the  world  "ready"  to  receive  it. 

In  handling  these  concepts  of  Zeitgeist,  configuration,  etc., 
it  is  difficult  to  define  their  essential  meaning  without  in- 
voking some  sort  of  mysticism.  Their  exponents  have  in 
general  taken  the  wiser  course,  illustrating  the  concepts  with 
concrete  examples  rather  than  giving  abstract  definitions  of 
the  terms  which  they  use,  and  it  is  unfortunate  that  even  this 

^  Benedict,  "Psychological  Types  in  the  Cultures  of  the  South-west", 
Proc.  22rd  Internal.  Congress  of  Americanists,  1928,  pp.  572-581 ;  Patterns 
of  Culture,  New  York,  1934.  I  have  been  much  indebted  to  the  ideas  of 
Dr  Benedict  and  at  the  end  of  this  book  an  attempt  will  be  made  to  indicate 
the  relations  between  my  concepts,  ethos  and  eidos,  and  her  concept  of 
configuration  and  the  terms  Apollonian  or  Dionysian,  realist  or  non- 
realist,  which  she  uses  to  describe  cultural  emphases. 


Problems  and  Methods  of  Approach  113 

course  has  led  to  their  being  branded  as  mystics.  There  are, 
however,  certain  generalisations  which  seem  to  apply  to  all 
these  concepts.  In  the  first  place  the  concepts  are  in  all  cases 
based  upon  an  holistic  rather  than  upon  a  crudely  analytic 
study  of  the  culture.  The  thesis  is  that  when  a  culture  is 
considered  as  a  whole  certain  emphases  emerge  built  up 
from  the  juxtaposition  of  the  diverse  traits  of  which  the 
culture  is  composed. 

If  we  examine  the  content  of  these  emphases  we  find  that 
they  are  conceived  to  be  either  systems  of  thought  or  scales 
of  values. 

But  the  two  words,  thought  and  value,  are  terms  which 
have  been  snatched  from  the  jargon  of  individual  psychology ; 
we  must  therefore  consider  in  what  sense  a  culture  may  be 
supposed  to  possess  either  a  system  of  thought  or  a  scale 
of  values.  At  the  present  time,  we  must  follow  the  opinion 
of  the  majority  of  psychologists  in  dismissing  the  theory  of 
the  group  mind  as  unnecessary,  and  therefore  regard  all  the 
thinking  and  feeling  which  occurs  in  a  culture  as  done  by 
individuals.  Thus  when  we  attribute  a  system  of  thought  or 
a  scale  of  values  to  a  culture,  we  must  mean  that  the  culture 
in  some  way  affects  the  psychology  of  the  individuals,  causing 
whole  groups  of  individuals  to  think  and  feel  alike. 

There  are  two  ways  in  which  culture  might  do  this,  either 
by  education,  inducing  and  promoting  certain  types  of  psycho- 
logical process,  or  by  selection,  favouring  those  individuals 
who  have  an  innate  tendency  to  psychological  processes  of  a 
certain  kind.  In  the  present  state  of  our  knowledge  of 
genetics,  we  cannot  pretend  to  estimate  the  relative  import- 
ance of  these  two  methods  of  changing  the  psychology  of  a 
population.  We  can  only  suppose  that  both  the  method  of 
selection  and  the  method  of  education  are  at  work  in  every 
community.  For  convenience,  I  shall  dodge  the  issue  of 
choosing  between  the  two  hypotheses  by  using  a  non- 
committal term  which  shall  subsume  them  both.  Following 
Dr  Benedict  I  shall  speak  of  culture  as  standardising  the 
psychology  of  the  individuals.  This  indeed  is  probably  one 
of  the  fundamental  axioms  of  the  holistic  approach  in  all 
sciences :  that  the  object  studied — be  it  an  animal,  a  plant  or 


114  Problems  and  Methods  of  Approach 

a  community — is  composed  of  units  whose  properties  are  in 
some  way  standardised  by  their  position  in  the  whole  organisa- 
tion. The  time  is  not  yet  ripe  for  any  detailed  analysis  of  the 
possible  standardising  effects  which  culture  may  have  upon 
the  individuals  in  the  community,  but  we  may  say  at  once 
that  culture  will  affect  their  scale  of  values.  It  will  affect 
the  manner  in  which  their  instincts  and  emotions  are  or- 
ganised into  sentiments  to  respond  differentially  to  the  various 
stimuli  of  life ;  we  may  find,  for  example,  that  in  one  culture 
physical  pain,  hunger,  poverty  and  asceticism  are  associated 
with  a  heightening  of  pride,  while  in  another,  pride  is  asso- 
ciated with  the  possession  of  property,  and  in  another  again, 
pride  may  be  even  gratified  by  public  ridicule. 

The  effects  of  the  culture  upon  the  system  of  thought  of 
the  individuals  are,  however,  not  so  clear.  That  the  circum- 
stances of  a  man's  life  will  affect  the  content  of  his  thought 
is  plain  enough,  but  the  whole  question  of  what  we  mean  by 
a  system  of  thought  remains  to  be  elucidated.  I  shall  there- 
fore leave  this  question  for  examination  in  a  later  chapter  and 
proceed  with  the  study  of  the  standardisation  of  the  affective 
aspects  of  the  individual's  psychology. 

Psychological  Theories  and  Ethology.  With  this  theory, 
that  a  culture  may  standardise  the  affective  make-up  of  indi- 
viduals, we  may  now  turn  to  the  theories  of  those  who  have 
sought  to  explain  social  phenomena  upon  psychological 
grounds.  These  theories  are  based  upon  broad  statements 
that  human  beings,  men  or  women  or  both,  in  all  races  and 
all  parts  of  the  world,  have  certain  fixed  patterns  of  emotional 
reaction.  In  applying  such  a  theory  to  our  naven  ceremonies 
we  might  for  example  say  that  men  have  naturally  certain 
attitudes  towards  women  and  that  therefore  whenever  men 
dress  as  women  their  behaviour  is  exaggerated  into  buf- 
foonery; while  women,  on  the  other  hand,  are  affected  in 
their  own  special  way  when  they  dress  up  as  men  and  there- 
fore they  put  on  a  prodigious  amount  of  swagger.  Or  again 
we  might  say  that  "human  beings  are  naturally  gregarious" 
and  that  this  fact  is  a  complete  and  sufficient  explanation 
of  the  large  size  of  the  latmul  village.  When  faced  with  the 
small  size  of  Mundugumor  villages  we  may  say  that  their 


Problems  and  Methods  of  Approach  115 

smallness  is  due  to  the  natural  hostility  which  exists  between 
males. 

When  stated  in  this  way  the  theories  have  a  slightly  ridicu- 
lous appearance,  but  it  is  worth  while  to  consider  the  position 
in  which  we  should  find  ourselves  if  we  indulged  the  facile 
building  of  these  theories  to  an  unlimited  extent.  We  should 
find  that  we  had  attributed  to  the  human  race  a  large  number 
of  conflicting  tendencies  and  that  we  had  invoked  certain 
tendencies  in  the  interpretation  of  one  culture  and  other, 
perhaps  opposite,  tendencies  in  the  interpretation  of  another. 
Such  a  position  is  untenable  unless  we  have  some  criterion 
whereby  we  may  justify  our  choice  of  a  particular  psycho- 
logical potentiality  for  use  in  interpreting  a  particular  cul- 
ture— some  criterion  whereby  we  may  decide  which  poten- 
tialities may  legitimately  be  invoked  in  describing  a  given 
culture.  But,  inasmuch  as  human  beings  often  appear  to 
harbour  conflicting  tendencies  and  potentialities,  this  position, 
with  all  its  contradictions,  may  become  tenable  as  soon  as  a 
satisfactory  criterion  is  discovered. 

Such  a  criterion  may,  I  believe,  be  derived  from  the  con- 
clusion at  which  we  arrived  above  in  our  examination  of  the 
concepts  of  the  historian.  We  concluded  that  culture  stan- 
dardises the  emotional  reactions  of  individuals,  and  modifies, 
the  organisation  of  their  sentiments;  in  fact,  that  culture 
modifies  the  very  same  aspects  of  the  individual  which  are 
invoked  by  the  rough  and  ready  psychological  theories  of 
culture.  We  must  therefore  re-phrase  the  psychological 
theories  in  some  such  terms  as  these :  A  human  being  is  born 
into  the  world  with  potentialities  and  tendencies  which  may 
be  developed  in  very  various  directions,  and  it  may  well  be 
that  different  individuals  have  different  potentialities.  The 
culture  into  which  an  individual  is  born  stresses  certain  of 
his  potentialities  and  suppresses  others,  and  it  acts  selectively, 
favouring  the  individuals  who  are  best  endowed  with  the 
potentialities  preferred  in  the  culture  and  discriminating 
against  those  with  alien  tendencies.  In  this  way  the  culture 
standardises  the  organisation  of  the  emotions  of  individuals. 

So  long  as  we  bear  in  mind  this  process  of  standardisation, 
we  may  safely  invoke,  in  order  to  explain  the  culture,  the 


1 16  Problems  and  Methods  of  Approach 

sentiments  of  the  individuals ;  but  we  must  always  verify  that 
the  sentiments  invoked  are  actually  those  which  are  fostered 
by  the  culture  in  question.  In  the  case  of  Mundugumor 
culture,  if  it  can  be  shown  that  the  hostility  between  indi- 
viduals is  an  aspect  of  human  nature  which  is  as  a  matter 
of  fact  stressed  by  the  culture,  then  we  shall  be  justified  in 
referring  to  this  hostility  as  a  factor  which  contributes  to  cause 
the  people  to  live  in  small  villages.  In  the  same  way,  if  it 
could  be  shown  that  latmul  culture  stressed  man's  gregarious 
tendencies,  we  should  be  justified  in  regarding  the  gregarious 
sentiment  or  instinct  as  important  in  the  moulding  of  the 
culture.  As  a  matter  of  fact,  however,  this  facet  of  human 
nature  is  not  specially  stressed  in  latmul  culture  and  this 
explanation  must  therefore  be  discarded. 

I  shall  show  later  that  pride  is  so  stressed  and  that  the  pride 
is  of  a  sort  to  be  gratified  by  the  big  ceremonial  houses  which 
require  organised  labour  on  a  large  scale,  by  great  ceremonies 
and  dances  which  require  many  performers  and  by  head- 
hunting which  prospers  when  the  village  is  strong  in  numbers. 
Thus  the  large  size  of  the  village  serves  an  important  function 
in  gratifying  pride — an  attribute  of  human  nature  which  is 
much  stressed  in  latmul  culture  and  to  which  therefore  we 
are  justified  in  referring. 

The  essence  of  the  method  is  then  that  we  first  determine 
the  system  of  sentiments  which  is  normal  to  the  culture  and 
emphasised  in  its  institutions ;  and  when  this  system  is  identi- 
fied we  are  justified  in  referring  to  it  as  a  factor  which  has 
been  active  in  shaping  the  institutions.  It  will  be  observed 
that  the  argument  is  circular. 

In  part  the  circularity  is  due  to  a  characteristic  of  all 
scientific  methods — the  fact  that  we  must  observe  a  number 
of  comparable  phenomena  before  we  can  make  any  theoretical 
statement  about  any  one  phenomenon.  But  the  circularity 
in  the  present  case  is  also  in  part  due  to  the  nature  of  the 
phenomena  which  we  are  studying. 

If  we  are  studying  jealousy  and  the  institutions  which 
regulate  sexual  life,  we  may  argue  both  that  the  institutions 
stress  jealousy  and  that  jealousy  has  shaped  the  institutions. 
It  would  seem  indeed  that  circularity  is  a  universal  property 


Problems  and  Methods  of  Approach  117 

of  functional  systems,  and  that  it  may  be  recognised  even  in 
such  crude  and  simple  systems  as  the  machines  devised  by 
man.  In  the  motor-car  for  example  the  magneto  produces 
electricity  because  the  engine  is  running  and  the  engine  runs 
because  of  the  sparks  provided  by  the  magneto.  Each  ele- 
ment in  the  functional  system  contributes  to  the  activity  of 
the  others  and  each  is  dependent  upon  the  activity  of  the 
others. 

As  long  as  we  take  an  external — behaviouristic — view  of 
a  functional  system  we  can  avoid  statements  of  circularity. 
We  can  see  a  motor-car  as  a  thing  into  which  petrol  is  poured 
and  which  runs  along  the  road  producing  smoke  and  killing 
pedestrians.  But  the  moment  we  turn  from  this  external  view 
and  begin  to  study  the  internal  workings  of  the  functional 
system  we  are  forced  to  accept  the  fundamental  circularity 
of  the  phenomena.  And  this  acceptance  is  demanded  not 
only  by  ethology  but  by  the  whole  functional  approach  to 
anthropology ;  and  the  students  who  are  engaged  in  working 
from  this  point  of  view  have  realised  this.  Thus  Malinowski 
claims  that "  the  functional  view  avoids  the  error  of  attributing 
priority  to  one  or  the  other  aspect  of  culture.  Material 
objects,  social  grouping,  traditional  and  moral  values,  as  well 
as  knowledge  are  all  welded  into  a  functional  system.  "^ 

A  further  and  more  compelling  argument  in  favour  of 
the  circular  or  reticulate  view  of  functional  systems  is  to  be 
found  in  the  fact  that  any  other  view  would  drive  us  to 
belief  either  in  a  "first  cause"  or  in  some  sort  of  teleology 
— in  fact  we  should  have  to  accept  some  fundamental 
dualism  in  nature  which  is  philosophically  inadmissible.^ 

Thus,  since  the  phenomena  which  we  are  studying  are 
themselves  inter-dependent  it  is  certain  that  our  descriptions 
must  contain  inter-dependent  statements;  and  since  this  is 
so  the  descriptions  must  for  ever  be  regarded  as  "not  proven " 
unless  we  can  devise  some  method  of  transcending  the  limits 
of  the  circles.  In  a  functional  analysis  we  subdivide  the 
systems  studied  into  a  series  of  parts  or  elements  and  produce 
theories  about  the  functional  relationships  between  these 

^  Encyclopaedia  Britannica,  art.  Anthropology. 

•  Cf.  Whdtehead,  The  Concept  of  Nature,  1920,  especially  chap.  II. 


1 1 8  Problems  and  Methods  of  Approach 

elements.  So  long  as  we  study  a  single  system  these  statements 
are  bound  to  be  circular  and  therefore  not  proven.  But  if  we 
could  extract  comparable  parts  from  different  systems  and 
verify  that  a  given  element  has  the  same  function  in  different 
systems  we  could  finally  verify  the  statements. 

The  orthodox  functional  school  has  adopted  the  practice 
of  dividing  cultures  into  institutions.  But  since  the  same 
institution  is  liable  to  have  the  most  various  functions  in 
different  societies  the  final  verification  of  the  theories  is  im- 
possible. If  we  take  the  institution  of  marriage  we  find  that 
it  may  function  variously  in  the  determination  of  status  of 
offspring,  in  the  regulation  of  sex  life,  in  the  education  of 
offspring,  in  the  regulation  of  economic  life,  etc.;  and  we 
find  that  the  relative  importance  of  these  functions  in  different 
cultures  varies  so  widely  that  it  is  almost  impossible  to  verify 
by  comparative  methods  the  truth  of  any  statement  which 
we  may  make  about  marriage  in  any  one  culture. 

The  ethological  approach  involves  a  very  different  system 
of  subdivision  of  culture.  Its  thesis  is  that  we  may  abstract 
from  a  culture  a  certain  systematic  aspect  called  ethos  which 
we  may  define  as  the  expression  of  a  culturally  standardised 
system  of  organisation  of  the  instincts  and  emotions  of  the 
individuals.  The  ethos  of  a  given  culture  is  as  we  shall  see 
an  abstraction  from  the  whole  mass  of  its  institutions  and 
formulations  and  it  might  therefore  be  expected  that  ethoses 
would  be  infinitely  various  from  culture  to  culture — as 
various  as  the  institutions  themselves.  Actually,  however,  it 
is  possible  that  in  this  infinite  variousness  it  is  the  content 
of  affective  life  which  alters  from  culture  to  culture,  while 
the  underlying  systems  or  ethoses  are  continually  repeating 
themselves.  It  seems  likely — a  more  definite  statement  would 
be  premature — that  we  may  ultimately  be  able  to  classify  the 
types  of  ethos. 

The  psychologists  are  already  at  work  on  the  grading  and 
classification  of  individuals  and  already  it  seems  certain  that 
different  types  of  individual  are  prone  to  different  systems  of 
organisation  of  their  emotions  and  instincts.  If  this  be  so, 
then  there  is  a  strong  probability  that  the  types  of  ethos  will 
fall  into  the  same  categories  of  classification  as  the  individuals 


Problems  and  Methods  of  Approach  1 19 

and  therefore  we  may  expect  to  find  similar  ethos  in  different 
cuhures  and  shall  be  able  to  verify  our  conclusions  as  to  the 
functional  effects  of  any  one  type  of  ethos  by  comparison  of 
its  expression  in  one  culture  with  its  expression  in  another, 
thus  finally  transcending  the  limits  of  the  circular  argu- 
ment,^ 


EXAMPLES  OF  ETHOS  IN  ENGLISH  CULTURE 

Before  describing  the  ethos  of  latmul  culture,  I  shall  illustrate 
the  ethological  approach  by  some  examples  taken  from  our 
own  culture  in  order  to  give  a  clearer  impression  of  what  I 
mean  by  ethos.  When  a  group  of  young  intellectual  English 
men  or  women  are  talking  and  joking  together  wittily  and 
with  a  touch  of  light  cynicism,  there  is  established  among 
them  for  the  time  being  a  definite  tone  of  appropriate 
behaviour.  Such  specific  tones  of  behaviour  are  in  all  cases 
indicative  of  an  ethos.  They  are  expressions  of  a  standardised 
system  of  emotional  attitudes.  In  this  case  the  men  have 
temporarily  adopted  a  definite  set  of  sentiments  towards  the 
rest  of  the  world,  a  definite  attitude  towards  reality,  and  they 
will  joke  about  subjects  which  at  another  time  they  would 
treat  with  seriousness.  If  one  of  the  men  suddenly  intrudes 
a  sincere  or  realist  remark  it  will  be  received  with  no  en- 
thusiasm— perhaps  with  a  moment's  silence  and  a  slight 
feeling  that  the  sincere  person  has  committed  a  solecism. 

^  In  a  discussion  of  this  matter  which  I  have  had  with  Professor  F.  C. 
Bartlett,  he  maintained  that  the  position  is  not  really  circular  but  "  spiral " ; 
and  that  therefore  this  attempt  both  to  justify  and  to  escape  from  the 
circularity  is  irrelevant.  But  he  is  also  sceptical  of  the^  validity  of  isolating 
sjTichronic  from  diachronic  enquiries  and  therefore  does  not  think  in 
purely  synchronic  terms.  The  isolation  is,  I  agree,  artificial,  but  I  think 
that  it  is  useful  and  necessary  in  the  present  state  of  our  science.  Later, 
when  we  know  more  about  the  diachronic  aspects  of  culture,  it  may  be 
possible  to  synthesise  the  two  methods  of  approach.  It  is  probable  that 
we  shall  then  see  that  what  appear  to  be  circles  in  a  synchronic  "  trans- 
verse section"  of  a  culture  are  really  spirals.  Such  a  synthesis  will  of 
course  let  us  out  of  the  unproven  circular  argument.  But  if  we  are  to 
escape  from  it  without  altering  our  synchronic  point  of  view,  this  escape 
must  be  by  way  of  the  comparative  method,  the  classification  of  the  various 
ethoses  and  comparison  of  the  ethoses  with  their  associated  cultural 
systems. 


I20  Problems  and  Methods  of  Approach 

On  another  occasion  the  same  group  of  persons  may  adopt^ 
a  different  ethos;  they  may  talk  realistically  and  sincerely. 
Then  if  the  blunderer  makes  a  flippant  joke  it  will  fall  flat 
and  feel  like  a  solecism. 

The  point  which  I  wish  to  stress  in  this  example  is  that 
any  group  of  people  may  establish  among  themselves  an 
ethos  which  as  soon  as  it  is  established  becomes  a  very  real 
factor  in  determining  their  conduct.  This  ethos  is  expressed 
in  the  tone  of  their  behaviour.  I  have  deliberately  for  my 
initial  example  chosen  an  instance  of  labile  and  temporary 
ethos  in  order  to  show  that  the  process  of  development  of 
ethos,  far  from  being  mysterious  and  rare,  is  an  everyday 
phenomenon.  The  same  group  of  intellectuals  were  at  one 
time  serious  and  at  another  witty,  and  if  the  blunderer  had 
had  sufficient  force  of  personality  he  could  have  swung  the 
group  from  one  ethos  to  the  other.  He  could  have  influenced 
the  evolution  of  ethos  within  the  group. 

But  if,  instead  of  such  a  temporary  conversation  group,  we 
examine  some  more  formed  and  permanent  group — say  an 
army  mess  or  a  college  high  table — whose  members  con- 
tinually meet  under  the  same  conditions,  we  find  the  etho- 
logical  position  much  more  stable.  In  the  more  casual  groups 
sometimes  one  sort  of  remark  and  sometimes  another  is 
inappropriate,  but  in  any  formed  group  we  find  certain  types 
of  remark,  certain  tones  of  conversation  permanently  taboo. 
The  ethoses  of  the  formed  groups  are  still  not  absolutely 
fixed.  The  processes  of  ethological  change  are  still  at  work 
and  if  we  could  compare  a  college  high  table  or  an  officers' 
mess  of  fifty  years  ago  with  those  groups  as  they  are  to-day 
we  should  no  doubt  find  very  considerable  changes.  Such 
changes  are  only  very  much  slower  in  the  formed  groups,  and 
enormously  greater  force  of  character  or  force  of  circum- 
stances is  required  suddenly  to  shift  the  ethos. 

Correlated  with  this  greater  stability  of  ethos,  there  is  a 
new  phenomenon  present  in  the  formed  groups  which  was 

^  The  factors — weather,  mood,  outside  events,  interactions  of  per- 
sonalities, etc. — which  cause  these  changes  of  ethos  have  never  been 
investigated.  Their  study,  though  laborious  and  difficult,  will  certainly 
throw  light  upon  many  of  our  problems. 


Problems  and  Methods  of  Approach  121 

absent  or  scarcely  recognisable  in  the  unformed.  The  group 
has  developed  its  own  cultural  structure  and  its  own  "tradi- 
tions" which  have  grown  up  hand  in  hand  with  the  ethos. 
At  the  high  table  (a  group  which  is  more  familiar  to  me  than 
is  the  army  mess)  we  find  such  cultural  developments  as  the 
Latin  Grace,  the  dons'  gowns  and  the  silver  presented  to  the 
College  by  former  generations  of  Fellows.  All  these  things 
have  their  effect  in  emphasising  and  stabilising  the  ethos  of 
the  group;  and  we  cannot  in  any  instance  say  that  a  given 
detail  is  due  exclusively  either  to  tradition  or  to  the  present 
ethos.  The  dons  of  St  John's  College  drink  water,  beer, 
claret,  sherry  and  port — but  not  cocktails ;  and  in  their  choice 
they  are  guided  both  by  tradition  and  by  the  ethos  of  the 
group.  These  two  factors  work  together  and  we  may  say  that 
the  dons  drink  as  they  do  both  because  generations  of  dons 
have  drunk  on  the  same  sound  system  in  the  past  and  because 
actually  in  the  present  that  system  seems  to  them  appropriate 
to  the  ethos  of  their  society.  Whatever  detail  of  the  tradition 
we  examine,  the  same  considerations  apply.  The  Latin  Grace, 
the  architecture  of  the  college,  the  snuff  after  dinner  on 
Sundays,  the  loving  cup,  the  rose  water,  the  feasts — all  these 
cultural  details  constitute  an  intricate  series  of  channels 
which  express  and  guide  the  ethos. ^  The  details  were  in  the 
past  selected  by  the  ethos  and  are  still  preserved  by  it.  The 
system  is  a  circular  one ;  and  the  very  attitude  which  the  dons 
adopt  towards  the  past  has  been  historically  formed  and  is 
an  expression  of  their  present  ethos. 

This  intimate  relationship  between  ethos  and  cultural  struc- 
ture is  especially  characteristic  of  small  segregated  groups 
where  the  ethos  is  uniform  and  the  "tradition"  very  much 
alive.  Indeed  when  we  say  that  a  tradition  is  "alive"  what 
we  mean  is  simply  this,  that  it  retains  its  connection  with  a 
persisting  ethos.  But  when  we  come  to  consider  not  isolated 
groups  but  whole  civilisations  we  must  expect  to  find  much 
more  variety  of  ethos  and  more  details  of  culture  which  have 

^  Such  metaphors  as  this  are  of  course  dangerous.  The  ir  use  encourages 
us  to  think  of  ethos  and  structure  as  different  "  things  "  instead  of  reahsing 
as  we  should  that  they  are  only  different  aspects  of  the  same  behaviour. 
I  have  let  the  metaphor  stand  pour  encourager  les  autres. 


122  Problems  and  Methods  of  Approach 

been  separated  from  the  ethological  contexts  in  which  they 
were  appropriate  and  retained  as  discrepant  elements  in  an 
otherwise  harmonious  culture.  Nevertheless  I  believe  that 
the  concept  of  ethos  may  valuably  be  applied  even  to  such 
enormous  and  confused  cultures  as  those  of  Western  Europe. 
In  such  cases  we  must  never  lose  sight  of  the  variations  of 
ethos  in  different  sections  of  the  community  and  the  curious 
dovetailing  of  the  ethoses  of  the  different  sections  into  an 
harmonious  whole,  whereby,  for  example,  peasants  with  one 
ethos  are  enabled  to  live  happily  under  feudal  lords  who  have 
a  different  ethos.  Differentiated  systems  of  this  kind  may 
persist  for  generations  and  only  break  down  when  the  scales 
of  values  are  questioned ;  when  the  lords  begin  to  doubt  the 
ethics  of  their  position  and  the  serfs  to  doubt  the  propriety 
of  submission — phenomena  which  are  liable  to  occur  when 
the  differentiation  has  proceeded  too  far. 


CHAPTER  IX 

The  Ethos  of  latmul  Culture:  the  Men 

COMPARED  with  European  ethology,  the  conditions 
among  the  latmul  are  remarkably  simple,  since 
their  culture  recognises  no  differentiation  of  rank  or 
class.  Indeed  the  only  social  differentiation  which  we  need 
to  consider  is  that  which  exists  between  the  sexes,  and,  since 
the  problems  which  we  are  studying  are  connected  with 
transvesticism,  it  is  the  differentiation  between  the  sexes 
which  is  most  likely  to  afford  clues. 

From  whatever  side  we  approach  the  culture,  whatever 
institutions  we  study,  we  find  the  same  sort  of  contrast 
between  the  life  of  the  men  and  that  of  the  women.  Broadly, 
we  may  say  that  the  men  are  occupied  with  the  spectacular, 
dramatic,  and  violent  activities  which  have  their  centre  in 
the  ceremonial  house,  while  the  women  are  occupied  with  the 
useful  and  necessary  routines  of  food-getting,  cooking,  and 
rearing  children — activities  which  centre  around  the  dwelling 
house  and  the  gardens.  The  contrast  between  ceremonial 
house  and  dwelling  house  is  fundamental  for  the  culture  and 
will  serve  as  the  best  starting-point  for  ethological  description. 
The  ceremonial  house  is  a  splendid  building,  as  much  as 
a  hundred  and  twenty  feet  in  length,  with  towering  gables 
^t'the  ends  (see  Plate  VII a).  Inside  the  building,  there 
is  a  long  vista  from  end  to  end  down  the  series  of  supporting 
posts  as  in  the  nave  of  a  darkened  church ;  and  the  resem- 
blance to  a  church  is  carried  further  in  the  native  attitudes 
towards  the  building.  There  is  a  series  of  taboos  on  any  sort 
of  desecration.  The  earth  floor  must  not  be  scratched  nor 
the  woodwork  damaged.  A  man  should  not  walk  right 
through  the  building  and  out  at  the  other  end;  he  should 
turn  aside  and  pass  out  by  one  of  the  side  entrances.  To  walk 
right  through  the  building  is  felt  to  be  an  expression  of  over- 
weening pride — as  if  a  man  should  lay  claim  to  the  whole 
building  as  his  personal  property. 


124        ^^^  Ethos  of  latmul  Culture:  the  Men 

But  the  analogy  between  ceremonial  house  and  church 
must  not  be  pushed  too  far  for  many  reasons :  the  ceremonial 
house  serves  not  only  as  a  place  of  ritual  but  also  as  a  club- 
house where  men  meet  and  gossip  and  as  an  assembly-room 
where  they  debate  and  brawl.  Further,  the  ceremonial  house 
does  not  stand  to  the  natives  as  a  symbol  of  their  devotion 
but  rather  as  a  symbol  of  their  pride  in  head-hunting.  Where 
we  think  of  a  church  as  sacred  and  cool,  they  think  of  a 
ceremonial  house  as  "  hot ",  imbued  with  heat  by  the  violence 
and  killing  which  were  necessary  for  its  building  and  conse- 
cration. Lastly,  the  ethos  of  behaviour  in  the  ceremonial 
house  is  as  far  removed  from  the  austerity  which  we  associate 
with  certain  churches  as  it  is  from  the  meek  devotion  asso- 
ciated with  others. 

Instead  of  austerity  or  meekness,  there  is  a  mixture  of 
pride  and  histrionic  self-consciousness. ^  An  important  man 
on  entering  the  ceremonial  house  is  conscious  that  the  public 
eye  is  upon  him  and  he  responds  to  this  stimulus  by  some 
sort  of  over-emphasis.  He  will  enter  with  a  gesture  and  call 
attention  to  his  presence  with  some  remark.  Sometimes  he 
will  tend  towards  a  harsh  sv/agger  and  over-consciousness 
of  pride,  and  sometimes  he  will  respond  with  buffoonery. 
But  in  whichever  direction  he  reacts,  the  reaction  is  theatrical 
and  superficial.  Either  pride  or  clowning  is  accepted  as 
respectable  and  normal  behaviour. 

In  this  community  there  are  no  steady  and  dignified  chiefs 
— indeed  no  formulated  chieftainship  at  all — but  instead 
there  is  continual  emphasis  on  self-assertion.  A  man  achieves 
standing  in  the  community  by  his  achievements  in  war,  by 
sorcery  and  esoteric  knowledge,  by  shamanism,  by  wealth, 
by  intrigue,  and,  to  some  extent,  by  age.  But  in  addition 
to  these  factors  he  gains  standing  by  playing  up  to  the  public 
eye;  and  the  more  standing  he  has,  the  more  conspicuous 

^  In  my  description  of  ethos,  I  have  not  hesitated  to  invoke  the  con- 
cepts of  emotion  and  to  use  terms  which  strictly  should  only  be  used  by 
observers  about  their  own  introspections.  I  have  been  driven  to  this  loose 
phrasing  through  lack  of  any  proper  technique  for  recording  and  of  any 
language  for  describing  human  gesture  and  behaviour.  But  I  wish  it  to 
be  understood  that  statements  of  this  kind  are  an  attempt — crude  and 
unscientific  perhaps — to  convey  to  the  reader  some  impression  of  the 
behaviour  of  latmul  natives. 


The  Ethos  of  latmul  Culture:  the  Men        125 

will  be  his  behaviour.  The  greatest  and  most  influential  men 
will  resort  freely  either  to  harsh  vituperation  or  to  buffoonery 
when  they  are  in  the  centre  of  the  stage,  and  reserve  their 
dignity  for  occasions  when  they  are  in  the  background. 

Among  the  younger  men  whose  standing  is  not  as  yet 
assured,  there  is  rather  more  self-control.  They  will  enter  the 
ceremonial  house  soberly  and  unobtrusively  and  sit  there 
quietly  and  gravely  in  the  presence  of  their  ranting  seniors. 
But  there  is  a  smaller  junior  ceremonial  house  for  the  boys. 
Here  they  carry  out  in  miniature  the  ceremonial  of  the  senior 
house  and  here  they  imitate  their  elders  in  mixing  pride  with 
buffoonery. 

We  may  summarise  the  ethos  of  the  ceremonial  house  by 
describing  the  institution  as  a  club — not  a  club  in  which  the 
members  are  at  their  ease,  but  a  club  in  which,  though 
separated  from  their  womenfolk,  they  are  acutely  conscious 
of  being  in  public.  This  self-consciousness  is  present  even 
at  times  when  there  is  no  specific  formal  or  ritual  activity 
taking  place,  but  it  is  enormously  more  marked  when  the 
men  are  assembled  in  the  ceremonial  house  for  some  debate 
or  ritual  performance. 

Any  matter  of  general  interest  may  be  disputed  formally 
in  a  traditional  manner.  In  every  large  ceremonial  house 
there  is  a  special  stool,  which  differs  from  the  ordinary  stools 
on  which  men  sit  in  having  a  "back",  like  a  chair,  carved 
into  some  representation  of  totemic  ancestors.  This  stool  is 
not  used  for  sitting  upon,  and  is  indeed  not  casually  touched 
if  it  be  an  old  and  sacred  specimen.^  It  is  used  solely  as  a 

*  As  far  as  I  could  make  out  the  degree  of  sacredness  of  the  debating 
stool  depends  simply  upon  its  age.  I  was  told  that  the  stool  acquires 
"heat"  through  the  rage  of  successive  debaters. 

In  the  account  which  I  have  published  of  latmul  debating  {OceaniOy 
1932,  p.  260)  a  slip  occurs.  I  there  state  "on  the  occasion  of  quiet  careful 
discussion  of  personal  names  and  genealogies  Dracaena  leaves  are  used, 
but  in  more  passionate  disputes,  bunches  of  coconut  leaflets  are  substi- 
tuted. These  latter  being  tougher  are  more  able  to  withstand  violent 
usage."  This  information  was  given  me  in  reply  to  a  question  as  to  the 
reasons  for  using  Dracaena  leaves.  I  had  only  seen  debates  in  which 
coconut  leaves  were  used,  but  when  visiting  the  ceremonial  house  in 
another  village  I  found  on  the  debating  stool  three  bunches  of  Dracaena 
leaves  and  therefore  asked  the  question.  Actually,  inasmuch  as  my  later 
observations  of  debates  about  names  and  ancestors  do  not  bear  out  the 
statement  that  such  debates  are  "quiet",  I  now  incline  to  the  view  that 


126        The  Ethos  of  lattnul  Culture:  the  Men 

table  for  debates.  The  speaker  has  three  bunches  of  Dracaena 
leaves,  or  coconut  leaflets.  He  picks  these  up  at  the  beginning 
of  his  speech  and  with  the  combined  bunches  he  gives  a  blow 
to  the  stool.  He  then  puts  down  the  bunches  on  the  stool, 
one  by  one,  as  if  they  were  a  tally  of  his  sentences.  When 
all  are  put  down,  he  again  bunches  them  together  and  gives 
another  single  blow.  This  series  of  actions  is  repeated  through- 
out his  speech,  ending  with  a  final  blow. 

The  tone  of  the  debates  is  noisy,  angry  and,  above  all, 
ironical.  The  speakers  work  themselves  up  to  a  high  pitch 
of  superficial  excitement,  all  the  time  tempering  their  violence 
with  histrionic  gesture  and  alternating  in  their  tone  between 
harshness  and  buffoonery.  The  style  of  the  oratory  varies  a 
good  deal  from  speaker  to  speaker  and  that  of  the  more 
admired  performers  may  tend  towards  the  display  of  erudi- 
tion or  towards  violence  or  to  a  mixture  of  these  attitudes. 
On  the  one  hand  there  are  men  who  carry  in  their  heads 
between  ten  and  twenty  thousand  polysyllabic  names,  men 
whose  erudition  in  the  totemic  system  is  a  matter  of  pride 
to  the  whole  village ;  and  on  the  other  hand  there  are  speakers 
who  rely  for  effect  upon  gesture  and  tone  rather  than  upon 
the  matter  of  their  discourse.  Such  a  man  will  make  a  speech 
in  which  there  is  only  the  barest  minimum  of  contribution 
to  the  issue — and  that  minimum  something  which  has  already 
been  said  by  other  speakers — but  he  will  fill  out  his  speech 
with  assertions  of  his  scorn  and  threats  that  he  will  rape  the 
members  of  the  opposition,  accompanying  the  word  with 
obscene  pantomimic  dance.  Meanwhile  the  insulted  will 
watch  and  smile  a  little,  or  laugh  aloud  and  shout  ironic 
encouragement  to  the  insulting  speaker.  Besides  these  two 
types — the  erudite  and  the  abusive — there  are  also  nervous 
and  apologetic  speakers  whose  contributions  to  the  debate 
are  despised.  These  men  generally  attempt  the  erudite  style, 
but  their  memories  are  undermined  by  nervousness  and  their 
"howlers"  are  laughed  at  by  the  audience. 

As  the  debate  proceeds,  both  sides  become  more  excited 
and  some  of  the  men  leap  to  their  feet,  dancing  with  their 

the  reason  given  for  the  use  of  Dracaena  leaves  was  an  incorrect  rationalisa- 
tion and  suspect  that  these  leaves  are  used  in  debates  about  ancestors 
because  of  the  ceremonial  associations  of  the  Dracaena. 


The  Ethos  of  latmul  Culture:  the  Men        127 

spears  in  their  hands  and  threatening  an  immediate  resort 
to  violence;  but  after  a  while  they  subside  and  the  debate 
goes  on.  This  dancing  may  occur  three  or  four  times  in  a 
single  debate  without  any  actual  brawling,  and  then  sud- 
denly some  exasperated  speaker  will  go  to  the  "root"  of  the 
matter  and  declaim  some  esoteric  secret  about  the  totemic 
ancestors  of  the  other  side,  miming  one  of  their  cherished 
myths  in  a  contemptuous  dance.  Before  his  pantomime  has 
finished  a  brawl  will  have  started  which  may  lead  to  serious 
injuries  and  be  followed  by  a  long  feud  of  killings  by  sorcery. 

The  emotions  which  are  so  dramatically  paraded  in  debate 
have  their  centre  in  pride,  and  especially  in  individualistic 
pride.  But  hand  in  hand  with  this,  there  is  developed  a 
prodigious  pride  in  the  totemic  ancestors  of  the  clan;  and 
most  of  the  debates  are  concerned  with  the  details  of  the 
totemic  system.  This  totemic  system  has  an  obvious  affective 
function — a  very  important  one  in  this  culture — of  providing 
the  members  of  every  clan  with  matter  for  self-congratula- 
tion. But  inversely  the  proud  ethos  of  the  culture  has  reacted 
in  a  curious  way  upon  the  system  and  though  we  are  not  here 
concerned  with  the  origins  of  latmul  totemism,  a  description 
of  the  system  is  relevant  as  indicating  the  emotional  back- 
ground against  which  we  are  to  see  the  naven  ceremonies. 

iThe  totemic  system  is  enormously  elaborated  into  a  series  \ 
of  personal  names,  so  that  every  individual  bears  names  of 
totemic  ancestors — spirits,  birds,  stars,  animals,  pots,  adzes, 
etc.  etc. — of  his  or  her  clan,  and  one  individual  may  have 
thirty  or  more  such  names.  Every  clan  has  hundreds  of  these 
polysyllabic  ancestral  names  which  refer  in  their  etymology 
to  secret  myths.  It  seems  that  the  effect  of  pride  upon  this 
system  has  been  to  corrupt  the  origin  myths,  so  that  to-day 
each  moiety  has  its  own  phrasing  of  the  origins  of  the  world 
according  to  which  that  moiety's  own  importance  is  stressed 
at  the  expense  of  the  other  moiety.  The  same  tendency 
extends  to  the  clans.  Groups  of  clans  flatter  themselves  by 
secretly  maintaining  that  they  are  not  really  members  of 
either  moiety  but  are  the  fom  et  origo  from  which  both 
moieties  sprang;  and  each  group  has  its  own  secret  mytho- 
logy to  support  its  secret  claims.  The  debates  about  totemism 
are  usually  concerned  with  attempts  to  steal  totemic  ancestors 


128        The  Ethos  of  latmul  Culture:  the  Men 

by  stealing  names  and  one  of  the  most  important  features 
of  every  latmul  ceremony  (except  initiation)  is  the  chanting 
of  name  songs,  whereby  the  members  of  the  clan  are  reminded 
of  the  importance  of  their  ancestors  and  the  system  is  con- 
tinually memorised. 

Actually,  as  a  result  of  the  overlapping  mythology  and  the 
stealing  of  names,  the  system  is  in  a  terribly  muddled  state. 
In  spite  of  this,  the  people  are  very  proud  not  only  of  the 
number  of  their  totemic  ancestors  and  their  esoteric  exploits 
in  the  origins  of  the  world  but  even  of  the  "  straightness  "  of 
their  song  cycles.  They  feel  that  the  whole  gigantic  system 
is  perfectly  schematic  and  coherent.  Thus  the  prevailing 
pride  which  has  led  them  to  build  up  this  mass  of  fraudulent 
heraldry  is  still  such  that  the  people  regard  the  resulting 
tangle  as  rigid  and  coherent. 

af  If  we  turn  to  the  ritual  connected  with  the  ceremonial 
house,  we  see  the  men,  as  a  group,  still  vying  with  each  other, 
but  in  spite  of  their  rivalry  managing  to  work  together  to 
produce  a  spectacle  which  the  women  shall  admire  and 
marvel  at.  Almost  without  exception  the  ceremonies  of  the 
men  are  of  this  nature ;  and  the  ceremonial  house  serves  as 
a  Green  Room  for  the  preparation  of  the  show.  The  men  put 
on  their  masks  and  their  ornaments  in  its  privacy  and  thence 
sally  forth  to  dance  and  perform  before  the  women  who  are 
assembled  on  the  banks  at  the  sides  of  the  dancing  ground. 
Even  such  purely  male  affairs  as  initiation  are  so  staged  that 
parts  of  the  ceremony  are  visible  to  the  women  who  form 
an  audience  and  who  can  hear  issuing  from  the  ceremonial 
house  the  mysterious  and  beautiful  sounds  made  by  the 
various  secret  musical  instruments — flutes,  gongs,  bull- 
roarers,  etc.  Inside,  behind  screens  or  in  the  upper  storey 
of  the  ceremonial  house,  the  men  who  are  producing  these 
sounds  are  exceedingly  conscious  of  that  unseen  audience 
of  women.  They  think  of  the  women  as  admiring  their  music, 
and  if  they  make  a  technical  blunder  in  the  performance,  it 
is  the  laughter  of  the  women  that  they  fear.^ 

^  For  a  sketch  of  the  technique  and  sociological  setting  of  flute  music 
in  latmul  culture,  cf.  The  Eagle,  St  John's  College  Magazine,  1935, 
vol.  XLViii,  pp.  158-170. 


The  Ethos  of  latmul  Culture:  the  Men        129 

The  same  emphasis  on  pride  occurs  in  many  other  contexts 
of  the  Hfe  of  the  men.  Here  it  must  be  remembered  that 
the  ceremonial  house  is  also  the  meeting  place  in  which  a 
great  deal  of  the  everyday  work  of  the  men's  lives  is  organised. 
In  it  their  hunting,  fishing,  building  and  canoe  cutting  are 
discussed;  and  these  activities  are  carried  out  in  the  same 
spectacular  manner  as  the  ritual.  The  men  form  parties  in 
their  big  canoes  to  go  fishing  or  hunting;  or  they  go  off 
together  in  big  groups  to  cut  trees  in  the  bush.  Such  parties 
are  called  together  by  rhythms  beaten  on  the  great  slit-gongs 
in  the  ceremonial  house ;  and  if  the  work  is  taking  place  close 
to  the  ceremonial  house  the  gongs  are  beaten  to  "put  life''^ 
into  the  workers.  Finally,  the  completion  of  every  consider- 
able task  is  marked  by  the  performance  of  some  spectacular 
dance  or  ceremony. 

Thus  it  comes  about  that  the  ritual  significance  of  the  ^ 
ceremonies  is  almost  completely  ignored  and  the  whole 
emphasis  is  laid  on  the  function  of  the  ceremony  as  a  means 
of  celebrating  some  labour  accomplished  and  stressing  the 
greatness  of  the  clan  ancestors.  A  ceremony  nominally  con- 
nected with  fertility  and  prosperity  was  celebrated  when  a 
new  floor  had  been  put  into  the  ceremonial  house.  On  this 
occasion  the  majority  of  informants  said  that  the  ceremony 
was  being  performed  "because  of  the  new  floor".  Only  a 
very  few  men  were  conscious  of,  or  interested  in,  the  ritual 
significance  of  the  ceremony;  and  even  these  few  were  in- 
terested not  in  the  magical  effects  of  the  ceremony  but  rather 
in  its  esoteric  totemic  origins — matters  of  great  importance 
to  clans  whose  pride  is  based  largely  upon  details  of  their 
totemic  ancestry.  So  the  whole  culture  is  moulded  by  the 
continual  emphasis  upon  the  spectacular,  and  by  the  pride 
of  the  male  ethos.  Each  man  of  spirit  struts  and  shouts,  play- 
acting to  convince  himself  and  others  of  the  reality  of  a  prestige 
which  in  this  culture  receives  but  little  formal  recognition. 

^  The  native  phrase,  which  is  here  translated  "  put  life  ",  is  yivut  taka-. 
yivut  is  the  ordinary  word  for  "movement"  or  "liveliness"  and  may  be 
used  as  noun,  adjective  or  verb;  e.g.  vavi  yivul  yi-rega-nda,  literally 
"bird  move  going-it",  i.e.  "the  bird  is  moving";  yivut  kami,  "a  live 
fish ".  taka-  is  the  ordinary  word  for  "to  put ",  " to  set ",  etc.  The  constant 
references  to  yivut  in  contexts  of  work,  play  and  ceremonial  should  be 
mentioned  as  characteristic  of  the  tempo  (cf.  p.  255)  of  the  culture. 


130       The  Ethos  of  latmul  Culture:  the  Men 

No  account  of  life  in  the  ceremonial  house  would  be  com- 
plete without  some  reference  to  the  ethos  of  initiation.  Here 
we  might  expect  in  another  culture  to  find  the  men  combining 
together  with  dignity  and  austerity  to  instruct  the  youths; 
and  in  the  painful  process  of  scarification  we  might  expect 
to  find  them  inculcating  Spartan  resistance  to  pain.  The 
culture  has  many  of  the  elements  which  would  seem  appro- 
priate to  such  an  ascetic  ethos:  there  are  days  during  which 
neither  novices  nor  initiators  may  eat  or  drink ;  and  there  are 
occasions  on  which  the  novice  is  made  to  drink  filthy  water. 
Again  the  culture  contains  elements  which  would  make  it 
appear  that  the  novice  is  passing  through  a  period  of  spiritual 
danger.  He  must  not  touch  his  food  with  his  hand  and  he  is 
subjected  to  a  drastic  washing  which  suggests  ritual  purifica- 
tion ;  and  so  we  might  reasonably  expect  to  see  the  initiators 
protecting  the  novices  from  dangerous  contamination. 

But  actually  the  spirit  in  which  the  ceremonies  are  carried 
out  is  neither  that  of  asceticism  nor  that  of  carefulness ;  it  is 
the  spirit  of  irresponsible  bullying  and  swagger.  In  the  pro- 
cess of  scarification  nobody  cares  how  the  little  boys  bear 
their  pain.  If  they  scream,  some  of  the  initiators  go  and 
hammer  on  the  gongs  to  drown  the  sound.  The  father  of  the 
little  boy  will  perhaps  stand  by  and  watch  the  process,  occa- 
sionally saying  in  a  conventional  way  "That's  enough!  that's 
enough ! "  but  no  attention  is  paid  to  him.  The  operators  are 
chiefly  interested  in  their  craft  and  regard  the  wriggling  and 
resistance  of  the  novice  as  prejudicial  to  it.  The  spectators 
are  rather  silent  with,  I  think,  a  touch  of  "cold  feet"  at  the 
sight  of  the  infliction  of  pain  divorced  from  the  normal 
setting  of  histrionic  excitement.    A  few  are  amused. 

When  pain  is  inflicted  in  other  parts  of  initiation,  it  is 
done  by  men  who  enjoy  doing  it  and  who  carry  out  their 
business  in  a  cynical,  practical-joking  spirit.  The  drinking 
of  filthy  water  is  a  great  joke  and  the  wretched  novices  are 
tricked  into  drinking  plenty  of  it.  On  another  occasion  their 
mouths  are  opened  with  a  piece  of  crocodile  bone  and  ex- 
amined "to  see  that  they  have  not  eaten  what  they  ought 
not".  They  are  not  under  any  food  taboos  at  this  time,  but 
the  result  of  the  examination  is  invariably  the  discovery  that 


The  Ethos  of  latmul  Culture:  the  Men  131 

the  mouth  is  unclean;  and  the  bone  is  suddenly  jabbed 
against  the  boy's  gums  making  them  bleed.  Then  the  process 
is  repeated  for  the  other  jaw.  In  the  ritual  washing,  the  partly 
healed  backs  of  the  novices  are  scrubbed,  and  they  are  splashed 
and  splashed  with  icy  water  till  they  are  whimpering  with 
cold  and  misery.  The  emphasis  is  upon  making  them  miser- 
able rather  than  clean. 

In  the  first  week  of  their  seclusion,  the  novices  are  sub- 
jected to  a  great  variety  of  cruel  and  harsh  tricks  of  this  kind 
and  for  every  trick  there  is  some  ritual  pretext.  And  it  is  still 
more  significant  of  the  ethos  of  the  culture  that  the  bullying 
of  the  novices  is  used  as  a  context  in  which  the  different 
groups  of  the  initiators  can  make  pride  points  against  each 
other.  One  moiety  of  the  initiators  decided  that  the  novices 
had  been  bullied  as  much  as  they  could  stand  and  were  for 
omitting  one  of  the  ritual  episodes.  The  other  moiety  then 
began  to  brag  that  the  lenient  ones  were  afraid  of  the  fine 
fashion  in  which  they  would  carry  out  the  bullying ;  and  the 
lenient  party  hardened  their  hearts  and  performed  the  episode 
with  some  extra  savagery. 

The  little  boy's  introduction  into  the  life  of  the  ceremonial 
house  is  conducted  on  such  lines  as  these  and  it  is  one  which 
fits  him  admirably  for  the  irresponsible  histrionic  pride  and 
buffoonery  which  is  characteristic  of  that  institution.  As  in 
other  cultures  a  boy  is  disciplined  so  that  he  may  be  able  to 
wield  authority,  so  on  the  Sepik  he  is  subjected  to  irrespon- 
sible bullying  and  ignominy  so  that  he  becomes  what  we 
should  describe  as  an  over-compensating,  harsh  man — whom 
the  natives  describe  as  a  "hot"  man. 

The  natives  themselves  have  summed  up  the  ethos  of 
initiation  in  a  formulation  which  is  especially  interesting  in 
view  of  the  contrasting  ethos  of  the  two  sexes.  During  the 
early  period  of  initiation  when  the  novices  are  being  merci- 
lessly bullied  and  hazed,  they  are  spoken  of  as  the  "wives" 
of  the  initiators,  whose  penes  they  are  made  to  handle.  Here 
it  seems  that  the  linguistic  usage  indicates  an  ethological 
analogy  between  the  relationship  of  man  and  wife  and  that 
of  initiator  and  novice.  Actually  resort  to  such  sadistic  treat- 
ment only  occurs  in  rather  extreme  circumstances  in  the 


132  The  Ethos  of  latmul  Culture:  the  Men 

case  of  wives,  but  it  is  perhaps  true  that  the  men  would  like 
to  believe  that  they  treat  their  wives  as  they  do  their  novices. 
I  think  we  may  see  a  consistent  cultural  pattern  running 
through  the  contrasting  sex  ethos,  the  shaming  of  the  novices, 
the  wau  shaming  himself  by  acting  as  the  wife  of  the  laua, 
and  the  use  of  the  exclamation  " Lan  men  to!"  (husband  thou 
indeed !),  to  express  contemptible  submission.  Each  of  these 
elements  of  culture  is  based  upon  the  basic  assumption  that 
the  passive  role  in  sex  is  shameful.^ 

In  fact,  the  initiatory  situation  is  not  a  simple  one  from  an 
ethological  point  of  view  but  is  essentially  a  contact  between 
two  ethoses,  that  of  the  initiators  and  that  of  the  novices. 
The  ethos  of  the  former  is  clearly  a  mere  exaggeration  of  that 
of  men  in  the  daily  life  of  the  ceremonial  house.  But  the 
ethos  of  the  novices  is  not  so  clear.  To  some  extent,  especially 
in  the  early  stages  of  initiation,  they  play  the  part  of  women ; 
and  we  may  ascribe  some  of  the  exaggeration  of  the  initiators* 
behaviour  to  the  presence  in  the  ceremonial  house  of  novices 
with  an  opposite  ethos. 

For  the  moment  we  are  not  concerned  with  how  this  con- 
trast has  arisen ;  but  we  may  suppose  that  the  little  boys  have 
to  some  extent  imbibed  the  women's  ethos  in  their  early 
life  and  so  come  to  their  initiation  with  some  of  the  emotional 
attitudes  characteristic  of  women  in  this  culture.  This  as- 
sumption I  cannot  definitely  state  to  be  founded  on  fact, 
because  I  did  not  study  the  children ;  but  whether  it  is  cor- 
rect or  not  it  is  certain  that  there  is  some  vague  idea  of  this 
sort  behind  the  initiation  rituals.  The  response  of  the  initiators 
to  this  real  or  nominal  contrast  between  the  novices  and  them- 
selves is  to  force  the  boys  further  into  the  complementary 
position,  dubbing  them  "wives"  and  bullying  them  into 
expressions  of  the  wifely  role. 

The  end  of  all  this  is  the  adoption  by  the  novices  of  the 
masculine  ethos,  but  it  seems  that  the  first  step  in  inducing 
this  process  is  to  compel  the  novices  to  behave  as  women, 
a  sufficiently  paradoxical  method  of  setting  about  the  busi- 
ness to  force  us  to  enquire  more  closely  into  the  processes 
which  it  involves. 

^  For  an  instance  of  the  opposite  opinion,  cf.  footnote,  p.  142. 


The  Ethos  of  latmul  Culture:  the  Men  133 

I  believe  that  we  may  distinguish^  four  contributory  pro- 
cesses : 

1.  A  process  whereby  the  novice  becomes  contra-sug- 
gestible to  the  female  ethos.  We  have  seen  that  the  treatment 
of  novices  qua  wives  is  a  great  deal  more  drastic  and  arbitrary 
than  the  treatment  of  real  wives ;  and  further  that  while  the 
women  grow  up  gradually  into  their  ethos,  the  novices  are 
suddenly  and  violently  plunged  into  submission.  These  dif- 
ferences account  I  believe  for  the  fact  that,  while  the  wives 
accept  a  somewhat  submissive  ethos  without  too  much  diffi- 
culty, the  novices  become  contra-suggestible  and  rebel  against 
it.  It  is  certain  that  revolts  and  refusals  to  undergo  further 
bullying  sometimes  occur  among  the  novices.  Such  revolts 
may  from  our  present  point  of  view  be  regarded  as  symptoms 
of  well  developed  contra-suggestibility  to  the  submissive  role, 
and  the  sharp  repressive  measures  to  which  the  initiators 
have  recourse  do  not,  we  may  suppose,  extinguish  the  resent- 
ment in  the  breasts  of  the  novices.^ 

2.  A  process  whereby  the  novices  become  proud  of  the 
male  ethos.  We  may  suppose  that  the  novices  derive  some 
feeling  of  superiority  from  their  separation  from  mothers  and 
sisters  and  especially  from  undergoing  an  experience  from 
which  women  and  smaller  children  are  excluded.  Certainly 
they  begin  at  quite  an  early  stage,  even  before  their  cuts  have 
healed,  to  be  proud  of  the  scars.  About  ten  days  after  the 
initial  scarring  the  novices  are  taken  out  into  the  bush  and 
are  threatened  with  bamboo  knives.  When  I  saw  this  done, 
four  out  of  the  five  novices  shrank  away  and  screamed  with 
something  approaching  hysterics  at  the  idea  of  further  cutting 
and  were  excused,  but  the  fifth  novice  after  a  few  moments' 
hesitation  submitted  to  the  operators  and  proudly  acquired 
extra  scars  without  flinching. 

^  The  theory  here  presented  to  a  considerable  extent  overweighs  the 
facts  \ipon  which  it  is  based.  This  analysis  of  the  initiatory  process  is, 
however,  intended  as  a  sample  of  the  ethological  point  of  view  and  an 
indication  of  the  sort  of  problems  which  this  point  of  view  raises,  not  as 
an  exposition  of  proven  hypothesis. 

^  Nowadays  it  is  often  necessary  to  initiate  boys  who  have  been  away 
to  work  for  the  white  man  and  who  return  as  grown  men  still  unscarred. 
Under  these  circumstances  such  revolts  are  especially  common  and  I  was 
twice  called  upon  by  the  initiators  to  help  in  the  preservation  of  the  system. 


134  The  Ethos  of  latmul  Culture:  the  Men 

After  the  first  week  of  intensive  bullying,  the  relationship 
between  initiators  and  novices  alters.  The  former  are  no 
longer  described  as  "elder  brothers"  or  "husbands"  but  are 
now  known  as  "mothers"  of  the  novices.  Corresponding  to 
this  change  of  phrasing,  the  initiators  turn  their  attention 
to  making  much  of  the  novices.  They  hunt  game  so  that  their 
"children"  may  grow  big  with  good  feeding,  they  teach 
them  how  to  play  the  flutes  and  they  undertake  as  communal 
tasks  the  making  of  various  ornaments — pubic  tassels,  lime 
boxes,  spear-throwers,  etc.  for  presentation  to  the  novices. 
Finally  at  the  end  of  initiation  the  novice,  decked  in  all  this 
finery,  is  exhibited  to  the  women  as  the  hero  of  the  occasion 
and  the  completion  of  his  initiation  is  celebrated  with  naven. 

3.  Reactions  to  the  presence  of  later  novices.  The  cere- 
monial end  of  his  initiation  by  no  means  completes  the 
assimilation  of  the  novice  into  the  group  of  the  initiators. 
There  are  other  younger  boys  who  though  too  young  for 
initiation  are  by  descent  of  the  same  grade  as  the  recently 
initiated  novices  (cf.  Fig.  4,  p.  245).  Later  when  these  boys 
come  up  for  initiation  the  former  novices  will  line  up  with 
them  to  be  shown  the  flutes  all  over  again.  They  will  not 
undergo  again  the  more  drastic  forms  of  bullying  but  cere- 
monially they  will  be  reckoned  as  novices.  At  all  these 
ceremonies  they  are  frankly  bored.  Only  much  later  when 
they  become  themselves  initiators  will  they  begin  to  take 
considerable  interest  in  the  business  of  initiation.  Then  they 
react  to  the  presence  of  novices  by  themselves  becoming 
bullies,  completely  assimilated  into  the  system. 

4.  Reactions  to  the  presence  of  other  initiators.  The 
drastic  bullying  behaviour  of  the  initiators  is  not  merely  a 
reaction  to  the  presence  of  novices,  but  is  also  promoted  to 
a  great  extent  by  the  feelings  of  emulation  which  exist  between 
members  of  the  rival  initiatory  groups.  While  group  Ax  3 
(cf.  Fig.  4,  p.  245)  is  initiating  By  4,  Ay  3  is  initiating  Bx  4 
on  the  opposite  side  of  the  ceremonial  house  and  between 
these  two  initiating  groups  there  is  constant  bickering  and 
rivalry  in  the  manner  of  their  performance.  On  one  occasion 
the  novices  had  been  so  severely  bullied  that  one  of  the 
bullying  tricks,  called  tshimangka,  the  fish,  in  which  the 


The  Ethos  of  latmul  Culture:  the  Men  135 

novices  would  be  very  severely  slapped,  was  postponed  from 
day  to  day.  Finally  the  time  drew  near  when  the  novices 
should  undergo  ritual  washing  after  which  the  bullying  would 
cease.  There  were  still  several  incidents  in  the  programme 
of  bullying  to  be  carried  out  and  the  principal  spokesman 
of  group  Ax  3  suggested  that  the  tshimangka  incident  should 
be  omitted.  The  immediate  retort  of  group  Ay  3  was  a 
boasting  taunt  to  those  who  suggested  leniency.  They  sug- 
gested that  Ax  3  were  afraid  of  the  brutal  way  in  which  they 
{Ay  3)  would  carry  out  the  rite.  In  the  face  of  this  accusation 
Ax  3  hardened  their  hearts.  The  tshimangka  was  immediately 
undertaken  and  carried  out.  When  it  was  over  and  while  the 
novices  were  spitting  the  blood  out  of  their  mouths,  the 
formerly  lenient  spokesman  assured  me  that  the  rite  had  been 
performed  with  extra  violence  in  reply  to  the  gibes  of  Ay  3.^ 

One  characteristic  of  the  latmul  men  is  very  clearly  ex- 
hibited in  the  contexts  provided  by  the  initiatory  system. 
This  is  their  tendency  to  "cut  off  their  own  noses  to  spite 
the  other  fellow's  face".  Whenever  a  serious  situation  arises 
in  the  initiatory  enclosure,  e.g.  when  a  woman  sees  something 
of  the  secrets,  or  some  disrespect  is  shown  to  the  secret 
objects,  or  when  some  serious  quarrel  breaks  out,  the  talk  is 
always  of  "breaking  the  screens",  throwing  the  whole  in- 
itiatory system  open  to  men,  women  and  children,  showing 
everything  to  everybody.  Much  of  this  talk  is  of  course  mere 
shouting  without  any  real  intention  of  drastic  action,  but 
from  time  to  time  an  impasse  causes  such  a  degree  of  exaspera- 
tion and  shame  that  the  men  carry  out  some  self-humiliating 
act  which  may  cripple  the  ceremonial  life  of  the  village  for 
some  years. 

Such  a  case  occurred  in  Mindimbit.  The  initiatory  system 
in  this  village  was  getting  feebler  and  feebler.  The  boys  were 
going  away  to  work  on  plantations,  leaving  the  village  too 
weak  in  numbers  for  any  great  ritual  to  be  attempted ;  while 
those  who  returned  to  the  village  did  so  with  a  scorn  of  the 

^  It  is  possible  that  a  fifth  factor — rivalry  between  villages — also  con- 
tributes to  shape  the  behaviour  of  the  initiators.  The  men  of  Komindimbit 
boasted  to  me  that  their  initiatory  crocodile  had  more  fierceness  {kau)  than 
that  of  any  other  village  on  the  river;  and  it  was,  I  think,  more  savage 
than  those  of  Palimbai  and  Kankanamun. 


136  The  Ethos  of  latmul  Culture:  the  Men 

initiatory  crocodile.  The  ceremonial  houses  were  full  of  boys 
with  no  scars  on  their  backs  and  one  small  ceremonial  house 
had  even  been  slowly  invaded  by  women  and  was  now 
deserted  by  the  men,  who  left  it  to  the  women  as  a  place 
where  they  might  sit  and  gossip. 

One  day,  a  party  of  Mindimbit  natives  were  being  given 
a  lift  on  a  white  recruiter's  schooner.  In  the  basket  of  one 
of  the  young  men  was  a  small  bamboo  end-flute.^  This 
object  was  noticed  by  the  man's  wife,  who  picked  it  up  and 
said  "What  is  this?"  An  older  man  was  present  and  saw 
the  incident.  He  scolded  the  woman  and  she  was  ashamed. 
Then  he  went  to  the  ceremonial  house  and  reported  what 
had  happened  and  abused  and  ranted  against  the  carelessness 
of  the  young  man  who  had  allowed  the  accident  to  occur. 
The  old  man  and  the  other  members  of  the  opposite  moiety 
raided  the  young  man's  house  and  smashed  his  wife's  pots. 
But  still  they  were  not  satisfied,  so  they  went  to  the  tagail 
or  junior  ceremonial  house^  and  to  the  dwelling  houses,  to 
collect  all  the  small  boys  together.  Even  the  small  toddlers 
who  could  only  just  talk  were  included.  Then,  that  night, 
they  showed  them  everything,  including  the  wagan  gongs. 
The  procedure  was  an  extreme  telescoping  of  the  whole 
initiatory  cycle,  but  was  unaccompanied  by  any  sort  of 
scarification.  This  drastic  event  was  regarded  both  in  Min- 
dimbit and  in  the  scornful  neighbouring  villages  as  the  final 
utter  shame  and  destruction  of  lesinduma,  the  village  "  croco- 
dile "  of  Mindimbit. 

Another  more  violent  but  similar  event  took  place  in 
Palimbai  some  fifty  years  ago.  The  village  was  celebrating 
its  wagan,  a  performance  on  the  gongs  very  much  more 

^  These  end-flutes  are  toys  made  by  work  boys  and  are  not  indigenous 
to  latmul  culture.  But  the  men  have  decided  that  all  wind  instruments 
ought  not  to  be  seen  by  women  lest  they  should  guess  at  the  nature  of 
the  flute  music  which  they  hear  coming  from  the  ceremonial  house. 

^  The  tagail  differs  from  the  mbicole  in  being  the  ceremonial  house  for 
uninitiated  members  of  jB  4  (cf.  Diagram,  p.  245),  while  the  mbwole 
belongs  to  A  $.  In  most  villages  only  one  of  these  buildings  is  erected 
and  serves  indiscriminately  for  both  purposes.  In  Mindimbit  there  was 
a  new  tagail  which  had  been  built  in  the  hope  that  its  presence  might 
attract  the  little  uninitiated  boys  away  from  the  senior  ceremonial  house. 
But  in  this  it  failed.  The  tagail  in  Mindimbit  was  generally  empty. 


The  Ethos  of  latmul  Culture:  the  Mm  137 

secret  and  serious  than  the  earher  initiation  ceremonies  which 
only  involve  flutes,  bullroarers,  etc.  In  the  wagan  ceremonies, 
the  secret  gongs  are  beaten  continuously, ^  day  and  night,  for 
months  on  end,  in  the  upper  storey  of  the  ceremonial  house ; 
and  during  all  this  time  there  must  be  no  noise  in  the  village : 
no  person  may  quarrel  or  shout  or  break  firewood.  The 
spears  stand,  ever  ready,  leaning  against  the  outside  of  the 
ceremonial  house  and  against  the  screens,  to  kill  any  man 
or  woman  who  offends  against  the  wagan  by  disturbing  the 
peace.  But  in  spite  of  this  preparedness  for  killing  the  spears 
are  not  often  used,  and  the  incident  which  occurred  in 
Palimbai  is  still  remembered  vividly  and  is  cited  as  proof 
that  the  spears  are  there  for  serious  use. 

The  ceremony  had  proceeded  without  serious  trouble  and 
it  was  time  to  prepare  for  the  final  spectacle  in  which  old 
men,  impersonating  the  wagan,  perform  a  dance  in  front  of 
the  assembled  women.  The  initiatory  group  who  were  pro- 
ducing the  spectacle  went  out  to  get  croton  {Codiaeum)  leaves 
with  which  to  ornament  the  giant  representations  of  wagan 
(cf.  Plate  XXVIII a),  sneaking  secretly  out  of  the  village  so 
that  no  women  should  know  of  the  methods  of  staging  the 
spectacle.  They  had  collected  the  leaves  and  put  them  in 
string  bags  and  were  on  their  way  back  to  the  village.  Some 
children  were  playing  near  the  mouth  of  the  waterway  which 
leads  in  from  the  river  to  Palimbai  village,  shooting  the 
straight  stalks  of  elephant  grass  with  toy  spear-throwers.  One 
of  these  missiles  fell  in  the  men's  canoe  and  pierced  the  string 
bag  in  which  the  croton  leaves  had  been  put. 

The  men  at  once  gave  chase  and  speared  the  small  boy  who 
was  guilty  of  this  offence.  When  they  got  back  to  the  village 
a  general  fight  ensued  in  which  three  (or  four?)  men  were 
killed,  all  of  them  members  of  the  clan  Wainggwonda, 
"fathers"  of  the  small  boy. 

^  This  is  a  statement  of  what  ought  to  happen.  Actually  the  rhythms 
are  rarely  maintained  for  more  than  two  or  three  days  and  are  generally 
interrupted  by  some  quarrelling  between  the  initiatory  moieties.  But  the 
rhythm  is  started  again  as  soon  as  each  quarrel  is  over  and  so  continues 
off  and  on  for  several  months.  When  one  performer  is  weary  another 
takes  his  place,  taking  the  moving  gong  stick  from  his  hand  so  that  not 
a  beat  is  missed. 


138  The  Ethos  of  latmul  Culture:  the  Men 

*Then  the  clan,  Tshimail,  which  had  taken  a  chief  part  in 
the  kiUing,  went  into  the  ceremonial  house  and  pulled  down 
the  wagan  gongs  from  the  upper  storey.  They  built  a  small 
screen  around  them  in  the  dancing  ground  and  then  showed 
the  gongs  to  all  the  women  of  Wainggwonda  and  handed  over 
the  sacred  gong  sticks  to  them,  to  keep  in  their  houses. 

In  killing  the  boy,  they  had  only  acted  according  to  the 
conventions  of  the  wagan  ceremony ;  but  still,  perhaps  feeling 
they  had  gone  too  far,  they  were  impelled  to  humiliate  them- 
selves, to  preserve  their  pride.  They  did  not  hang  tambointsha 
(tassels)  on  their  lime  sticks  for  these  killings. 

In  the  business  of  head-hunting,  the  masculine  ethos  no 
doubt  reached  its  most  complete  expression;  and  though  at 
the  present  time  the  ethos  of  head-hunting  cannot  be  satis- 
factorily observed,  there  is  enough  left  of  the  old  system  to 
give  the  investigator  some  impression  of  what  that  system 
implied.  Lacking  observations  of  actual  behaviour  my 
description  must,  however,  be  based  on  native  accounts. 

The  emphasis  here  was  not  on  courage ;  no  better  coup  was 
scored  for  a  kill  which  had  entailed  special  hardship  or 
bravery.  It  was  as  good  to  kill  a  woman  as  a  man,  and  as 
good  to  kill  by  stealth  as  in  open  fight.  An  example  will 
serve  to  illustrate  this  set  of  attitudes:  In  a  raid  on  one  of 
the  neighbouring  bush  villages  a  woman  was  killed  and  her 
daughter  was  taken  by  the  killer  (Malikindjin)  and  brought 
back  to  Kankanamun.  He  took  her  to  his  house,  where  for 
a  while  he  hid  her,  thinking  of  adopting  her  into  his  house- 
hold. But  she  did  not  remain  there.  He  took  her  to  the 
ceremonial  house  and  a  discussion  arose  as  to  her  fate.  She 
pleaded  that  she  should  be  pitied :  "You  are  not  my  enemies ; 
you  should  pity  me;  later  I  will  marry  in  this  village." 

One  of  the  young  men,  Avuran-mali,  son  of  her  captor,  cut 
into  this  discussion,  and  in  a  friendly  way  invited  her  to  come 
down  to  the  gardens  to  get  some  sugar-cane.  Accordingly 
he  and  the  girl  went  down  to  the  gardens  together  with  one 
or  two  of  the  younger  boys,  among  them  my  informant, 
Tshava,  who  was  then  a  small  boy.  On  arriving  there, 
Avuran-mali  speared  her.  (The  duty  of  cleaning  the  skull  fell 
to  Tshava.    An  enemy  skull  must  never  be  touched,  and 


The  Ethos  of  latmul  Culture:  the  Men  139 

Tshava  had  some  difficulty  in  detaching  a  Hgament.  He 
therefore  discarded  the  tongs,  seized  the  end  of  the  ligament 
in  his  teeth  and  pulled  at  it.  His  father  saw  him  and  was  very 
shocked;  but  Tshava  said  to  me:  "The  silly  old  man!  How 
was  I  to  know?" — an  attitude  towards  taboos  which  is  not 
uncommon  among  the  latmul.) 

But  in  spite  of  the  lack  of  "sportsmanship",  the  activity 
of  head-hunting  was  to  a  considerable  extent  a  "sport". 
There  was  no  clear  rule  that  you  must  have  a  grudge  against 
a  man  before  you  killed  him,  nor  against  a  village  before 
raiding  it;  though  a  majority  of  the  killings  were  certainly 
regarded  as  vengeance.  In  general  the  fighting  and  killing 
was  confined  to  the  killing  of  foreigners,  i.e.  members  of 
other  villages,  especially  of  villages  against  whom  a  feud 
existed.  But  even  this  rule  was  not  too  strictly  interpreted; 
a  woman,  married  into  the  village,  might  for  purposes  of 
head-hunting  be  considered  a  foreigner.  I  even  came  across 
one  case  in  which  a  man  wore  a  tassel  for  killing  his  own  wife 
in  revenge  for  a  kill  accomplished  by  members  of  the  village 
from  which  she  had  come. 

Two  main  motives  informed  this  system,  the  personal  pride 
of  the  individual  and  his  pride  and  satisfaction  in  the  pro- 
sperity and  strength  of  his  community.  These  two  motives 
were  closely  tangled  together.  On  the  purely  personal  side, 
the  successful  homicide  was  entitled  to  special  ornaments 
and  paints  and  to  the  wearing  of  a  flying  fox  skin  as  a  pubic 
apron ;  while  the  apron  of  stripped  Dracaena  leaves  was  the 
reproach  of  the  man  who  had  never  killed.  The  homicide 
was  the  hero  of  the  most  elaborate  naven  and  the  proud  giver 
of  feasts  to  his  lanoa  nampa  (husband  people).  Lastly,  he 
was  admired  by  the  women;  and  even  to-day  the  women 
occasionally  make  scornful  remarks  about  the  calico  loin- 
cloths worn  by  the  young  men  who  should  strictly  still  be 
wearing  Dracaena  aprons  like  those  which  were  given  them 
when  they  were  little  boys  being  initiated. 

The  association  of  personal  pride  with  success  in  head- 
hunting and  of  shame  with  failure  is  brought  out  too  in  the 
behaviour  of  those  whose  relatives  had  been  killed.  Their  first 
duty  was  the  taking  of  nggambwa  (vengeance).   The  rings  of 


140  The  Ethos  of  latmul  Culture:  the  Men 

cane  worn  in  mourning  for  the  killed  individual  may  not  be 
put  aside  until  vengeance  has  been  achieved;  and  a  pointed 
reference  to  an  unavenged  relative  is  one  of  the  most  dan- 
gerous insults  that  one  latmul  can  use  in  ranting  against 
another — an  insult  which  is  felt  to  be  specially  aggravating 
now  that  head-hunting  is  forbidden. 

Indeed  so  serious  is  the  condition  of  those  who  are  unable 
to  secure  revenge,  that  it  produces  ngglambi  in  the  group  and 
may  lead  to  the  sickness  and  death  of  its  members. 

This  spreading  of  clan  dysphoria  resulting  from  the  un- 
avenged insult  to  the  pride  of  the  clan  may  be  contrasted 
with  the  "sociological"  phrasings  of  the  benefits  which  suc- 
cessful head-hunting  confers  upon  the  community.  Here, 
as  is  usual  in  sociological  phrasings,  the  matter  is  expressed 
in  tangled  symbolism,  but  may  be  made  clear  by  an  artificial 
separation  of  the  various  components  of  the  system : 

1.  The  enemy  body  was,  if  possible,  brought  back  to  the 
village  and  was  there  ritually  killed  by  a  man  wearing  a  mask 
which  represents  an  eagle.  Thus  the  kill  symbolically  became 
the  achievement  not  only  of  the  individual  homicide  but  of 
the  whole  village. 

2.  The  natives  articulately  say  that  the  eagle  is  the  kau 
of  the  village.  Kau  is  a  word  which  means  "a  raiding  party  ", 
"a  fighting  force",  "an  expression  of  anger",  etc.  The  eagle 
is  also  represented  on  the  finial  of  the  ceremonial  house  (cf. 
Oceania,  1932,  Plate  VIII),  and  at  the  ceremony  with  which 
this  eagle  is  put  in  place,  the  bird  speaks.  He  looks  out  over 
the  enemy  country  and  sees  them  there  as  "birds  preening 
themselves"  or  as  "fish  jumping  in  the  water" — ready  to 
be  killed. 

3.  The  natives  say  that  prosperity — plenty  of  children, 
health,  dances  and  fine  ceremonial  houses — follows  upon 
successful  head-hunting. 

4.  Prosperity  is  also  dependent  upon  the  mbwan,  those 
ancestral^  spirits  which  are  represented  by  standing  stones. 

^  The  mbwan  are  regarded  as  ancestors  and  are  classified  roughly  with 
the  angk-au  or  potsherd  spirits.  But  in  some  cases  at  least,  the  mbwan 
are  really  the  spirits  not  of  deceased  ancestors  but  of  killed  enemies. 
Perhaps  they  are  thought  of  as  ancestors  because  of  their  activity  in 
promoting  the  proliferation  of  the  community. 


The  Ethos  of  latmul  Culture:  the  Men  141 

5.  The  heads  of  the  killed  were  placed  upon  the  mhwan 
and  in  some  cases  their  bodies  were  buried  under  the  mbwan. 

6.  The  standing  stones  are  phallic  symbols,  e.g.  in  the 
shaman's  jargon  the  phrase  for  copulation  is  mbwan  tou-, 
"setting  up  a  standing  stone". 

7.  The  male  sexual  act  is  definitely  associated  with  violence 
and  pride. 

Running  through  this  plexus  of  cultural  details  we  can 
clearly  see  the  general  position  of  head-hunting  as  the  main 
source  of  the  pride  of  the  village,  while  associated  with  the 
pride  is  prosperity,  fertility  and  the  male  sexual  act;  while 
on  the  opposite  side  of  the  picture  but  still  a  part  of  the 
same  ethos,  we  can  see  the  association  of  shame,  mourning 
and  ngglambi. 

Closely  linked  with  these  emphases  upon  pride  and  shame 
is  the  development  of  the  spectacular  side  of  head-hunting. 
Every  victory  was  celebrated  by  great  dances  and  ceremonial 
which  involved  the  whole  village.  The  killer  was  the  hero 
of  these  and  he  was  at  the  same  time  the  host  at  the  feasts 
which  accompany  them.  Even  the  vanquished  assented  to 
the  beauty  of  the  dances,  as  appears  from  a  text  collected  in 
Mindimbit  describing  the  typical  series  of  events  on  a  raid : 

(After  the  fighting)  they  leave  off.  Then  he  (the  killer, 
standing  in  his  canoe  and  holding  up  the  head  which  he  has 
taken  from  the  enemy)  asks  "I  am  going  to  my  beautiful^ 
dances,  to  my  beautiful  ceremonies.  Call  his  name."  (The 
vanquished  reply)  "It  is  so-and-so  that  you  have  speared." 
(Or  the  victor  will  say)  "  This  one  is  a  woman"  and  they  (the 
vanquished)  will  call  her  name  (and  they  will  cry  to  the 
victors)  "  Go.  Go  to  your  beautiful  dances,  to  your  beautiful 
ceremonies." 

^  In  this  text  the  phrase  which  I  have  translated  as  "beautiful  dances, 
beautiful  ceremonies "  is  of  considerable  interest.  The  native  word  for 
"beautiful"  is  yigen,  a  common  latmul  word  which  is  used  to  describe 
an  admired  face  or  spectacle.  The  same  word  also  occurs  in  the  adverb 
yigen-mbwa,  "gently",  the  opposite  of  nemwan-pa,  "violently"  (literally, 
"greatly").  The  whole  phrase  is  yigen  vi,  yigen  mbwanggo,  a  poetical 
form  built  up  out  of  the  common  everyday  phrase  vi  mbwanggo,  "a 
(triumphant)  war  dance".  In  this  phrase,  vi  is  the  word  for  a  particular 
sort  of  spear  with  many  points,  used  in  warfare,  and  mbwanggo  is  the 
ordinary  word  for  any  dance  or  ceremony.  In  the  traditional  diction  this 
phrase  is  divided  into  two  parallel  phrases,  a  common  trick  of  latmul 
poetic  genius.   (Cf.  also  yigen  kundi,  "quiet  singing",  p.  156.) 


CHAPTER  X 

The  Ethos  of  latmul  Culture:  the  Women 

IN  the  everyday  life  of  the  women  there  is  no  such  em- 
phasis on  pride  and  spectacular  appearance.^  The  greater 
part  of  their  time  is  spent  on  the  necessary  economic 
labours  connected  with  the  dwelling  house — food-getting, 
cooking  and  attention  to  babies — and  these  activities  are  not 
carried  out  publicly  and  in  big  groups,  but  privately  and 
quietly.  In  the  very  early  morning  before  dawn  the  women 
go  out  in  their  tiny  canoes  to  tend  the  fish  traps  in  which  they 
catch  the  prawns,  eels  and  small  fish  which  form  the  staple 
supply  of  protein  food.  Each  canoe  is  just  big  enough  to 
carry  a  woman  and  perhaps  her  small  child,  and  on  the  stern 
is  a  little  fire  in  an  old  pot  for  the  woman  to  warm  herself 
by  in  the  chilly  dawn ;  for  the  business  of  examining  the  traps 
involves  her  wading  about  in  water  up  to  her  breast  and  she 
will  be  cold  when  the  job  is  done.  The  little  fleet  of  canoes, 

^  There  is  some  local  difference  in  ethos  between  the  Eastern  latmul 
(Mindimbit,  Tambunum,  etc.)  and  the  Central  latmul  (Palimbai,  Kan- 
kanamun,  etc.).  Among  the  Eastern  people  the  women  wear  large  quanti- 
ties of  shell  ornaments  in  their  daily  life,  only  removing  them  for  such 
tasks  as  the  tending  of  fish  traps.  These  women  have  also  a  slightly 
prouder  bearing  than  the  women  of  Palimbai  and  Kankanamun,  who 
normally  wear  very  little  ornament.  This  difference  has  probably  some 
bearing  upon  the  culture  as  a  whole,  and  it  is  worth  mentioning  that 
among  the  Eastern  group  iai  marriage,  in  which  the  initiative  rests  with 
the  woman,  is  commoner  than  in  Palimbai ;  and  that  it  was  in  Mindimbit 
that  I  was  shown  a  very  interesting  lime  gourd.  It  is  the  custom  there 
for  men  to  scratch  on  their  gourds  tallies  of  their  successful  love  affairs ; 
but  the  gourd  in  question  besides  the  ordinary  tally  had,  incised  upon  it, 
a  large  representation  of  a  vulva  ornamented  with  geometrical  designs. 
I  asked  whether  it  was  carved  there  as  an  emblem  of  the  mother  moiety, 
but  the  owner  replied  with  pride :  "  No,  that  refers  to  a  woman.  I  did 
not  want  her,  but  she  came  to  my  mosquito  bag  and  took  the  active  role 
in  sex." 

In  general  there  is  the  same  sort  of  contrast  between  the  ethos  of  the 
two  sexes  in  both  areas,  but  this  contrast  is  most  marked  in  Kankanamun, 
where  the  women  are  definitely  shabby.  They  are  a  little  smarter  in 
Palimbai  and  markedly  smarter  in  Mindimbit.  The  men  of  Palimbai  are 
conscious  of  the  difference  between  their  women  and  those  of  Kankana- 
mun, and  attribute  it  to  the  better  supply  of  fish  which  they  get  because 
their  village  stands  on  the  banks  of  a  lake. 


The  Ethos  of  latmul  Culture:  the  Women      143 

each  with  its  column  of  smoke  rising  in  the  half  light,  is  a 
very  pretty  sight ;  but  as  they  draw  away  from  the  village  each 
canoe  separates  from  the  others  as  each  woman  goes  to  the 
part  of  the  river  where  her  traps  are  set.  In  this  work  there 
is  none  of  the  excitement  which  the  men  introduce  into  their 
fishing  expeditions.  Each  woman  goes  off  by  herself  to  do 
her  day's  work.  When  she  has  tended  the  fish  traps  she  will 
go  and  collect,  for  firewood,  old  dead  stems  of  the  elephant 
grass  which  lines  the  banks  of  the  river.  Then  she  will  return 
to  the  village  where  she  will  attend  to  the  cooking  and  the 
jobs  of  the  house. 

p  A  single  house  is  divided  between  two  or  three  men  related 
.  by  patrilineal  ties ;  and  this  division  of  the  house  is  felt  by 
the  men  to  be  very  real,  almost  a  matter  for  stiffness  and 
formality.  The  man  who  owns  one  end  of  the  house  will  avoid 
intruding  upon  his  brother's  or  his  son's  residence  at  the 
other  end^  although  there  is  no  screen  or  wall  dividing  the 
house,  only  the  big  sleeping  bags  in  the  centre  of  the  floor. 
But  although  typically  the  women  of  a  house  are  not  mutually 
related,  they  seem  to  be  much  less  conscious  of  the  divisions 
and  will  constantly  bandy  remarks  the  whole  length  of  the 
house.  Each  woman  has  her  separate  cooking  place  with  its 
fire  basins  set  up  close  to  the  wall  and  the  different  wives 
of  one  man  carry  out  their  cooking  independently,  but  still 
there  is  more  ease  in  their  mutual  relations  and  less  self- 
consciousness  than  is  the  case  among  the  men. 

The  women's  life  is  regulated  by  a  three-day  week,  the 
middle  day  of  each  three  being  a  market  day.^  The  supply 
of  fish  and  prawns  is  such  that  a  quantity  can  be  set  aside, 
either  kept  alive  in  baskets  immersed  in  the  water  or  smoked. 
On  a  market  day  the  women  gather  up  these  supplies  and 
go  off  in  their  canoes  to  the  bush  villages  where  they  barter 

^  It  is  my  impression  that  a  son  will  visit  his  father  much  more  freely 
and  casually  than  the  father  will  visit  the  son;  and  it  is  likely  that  there 
is  some  trace  of  the  same  tendency  between  brothers,  the  younger  visiting 
the  elder  more  freely. 

^  This  three-day  week  is  characteristic  of  Palimbai  and  Kankanamun, 
but  is  not  adhered  to  by  the  Eastern  latmul  whose  villages  are  farther 
from  those  of  their  bush  neighbours.  The  Eastern  latmul  hold  their 
markets  irregularly,  on  specially  arranged  days,  the  two  parties  meeting 
in  the  grass  country  which  separates  their  villages. 


144      ^^^  Ethos  of  latmul  Culture:  the  Women 

the  fish  for  sago.  The  market  may  be  held  actually  in  the 
bush  village,  but  more  often  the  women  of  the  bush  tribes 
come  half-way  along  the  road  and  the  parties  meet  at  some 
agreed  spot.  The  chaffering  is  done  easily  and  with  a  good 
deal  of  jolliness.  The  deals  are  small  and  there  is  very  little 
haggling  but  a  great  deal  of  talk,  not  only  about  the  matter 
in  hand  but  about  the  events  of  the  last  few  days  in  the 
various  villages.  At  these  markets  men  may  be  present,  but 
so  far  as  I  know  their  presence  has  no  quelling  effect  on  the 
general  ease  of  the  women.  I  have,  however,  only  attended 
markets  at  which  men  were  present  and  so  cannot  state 
definitely  what  effect  their  presence  has  upon  the  women. 
The  women's  markets  contrast  sharply  with  the  behaviour 
of  the  men  when  they  are  engaged  in  their  more  serious 
negotiations.  In  buying  a  sleeping  bag  or  a  canoe,  each  party 
tries  to  outdo  the  other  in  a  pose  of  critical  taciturnity,  and 
in  the  majority  of  cases  no  business  will  be  done. 

Compared  with  the  proud  men,  the  women  are  unostenta- 
tious. They  are  jolly  and  readily  co-operative  while  the  men 
are  so  obsessed  with  points  of  pride  that  co-operation  is 
rendered  difficult.  But  it  must  not  be  supposed  that  the 
women  are  mere  submissive  mice.  A  woman  should  know 
her  own  mind  and  be  prepared  to  assert  herself,  even  to  take 
the  initiative  in  love  affairs.  In  iai  marriage  (cf.  pp.  88, 89)  it  is 
nominally  the  woman  who  makes  the  advances  and  who,  of 
her  own  accord  and  uninvited,  goes  to  the  house  of  her 
ianan.  It  is  said  in  Tambunum  that  in  such  cases  the  chosen 
man  has  no  right  to  refuse  such  a  proposal. 

The  same  pattern  is  often  followed  in  less  formalised 
marriages  in  which  the  woman  has  no  such  nominal  right. 
A  typical  case  will  illustrate  the  extent  of  the  woman's 
initiative:  I  had  gone  with  my  native  servants  to  look  at 
some  ceremonies  in  the  neighbouring  village  of  Aibom,  a 
village  of  foreigners  who  are  not  regarded  as  true  latmul 
but  whose  social  system  is  very  closely  related  to  that  of  the 
latmul.  On  the  day  after  my  return  from  this  expedition, 
a  girl  from  Aibom  arrived  alone  in  Kankanamun.  She  en- 
quired for  members  of  her  own  clan  and  went  to  the  house 
of  one  of  my  informants  who  was  her  clan  brother.   She  told 


The  Ethos  of  latmul  Culture:  the  Women      145 

him  that  she  was  in  love  with  one  of  my  cook-boys.  Her  clan 
brother  asked  "Which  cook-boy?"  She  explained  that  she 
did  not  know  his  name.  So  they  went  together  to  a  spot 
where  they  could  observe  my  domestic  staff  and  the  girl 
pointed  out  the  cook-boy  whom  she  loved  and  he  was  thereby 
identified.  In  the  negotiations  which  followed  the  boy  and 
girl  modestly  avoided  each  other  (so  the  cook-boy  told  me) 
but  her  clan  brother  acted  as  intermediary  and  she  lodged 
in  his  house.  The  boy  was  definitely  flattered  by  the  proposal 
and  decided  to  accept  it.  He  sent  a  series  of  small  presents 
to  the  girl,  which  she  accepted.  Very  soon  messages  began 
to  come  from  Aibom  demanding  the  bride  price;  and  the 
cook-boy  was  not  rich.  There  were  delays,  and  after  about 
a  fortnight  the  girl  returned  to  Aibom.  The  boy  then  put  on 
some  show  of  anger  and  demanded  from  her  relatives  some 
return  for  the  presents  which  he  had  given  her.  But  this 
return  was  not  forthcoming. 

The  point  which  I  wish  to  stress  in  this  incident  is  the 
extraordinary  courage  which  the  girl  showed  in  coming  alone 
to  a  foreign  village,  and  the  clarity  with  which  she  knew  her 
own  mind.  Her  conduct  was  regarded  as  culturally  normal 
by  the  latmul. 

As  a  further  documentation  of  the  respect  which  is  paid 
to  women  of  strong  and  courageous  personality,  we  may  cite 
here  a  traditional  myth  which  was  told  to  me  in  Mindimbit 
in  explanation  of  the  head-hunting  alliance  between  that 
village  and  Palimbai.  Both  of  these  villages  have  a  tradi- 
tional feud  with  the  village  of  Kararau  which  lies  between 
them. 

Kararau  were  killing  us.  They  speared  women  who  went  out 
to  get  tips  of  wild  sugar-cane,  and  women  who  went  to  get  water- 
weed  (for  pigs'  food),  and  women  who  went  to  their  fish  traps. 
And  they  shot  a  man,  Au-vitkai-mali.  His  wife  was  Tshanggi-mbo 
and  (his  sister  was)  Au-vitkai-mangka.  They  shot  him  and  beat 
the  gongs  (in  triumph).  Au-vitkai-mangka  was  away;  she  was  on 
the  lake  (fishing).  Au-vitkai-mali  went  to  his  garden  and  they  speared 
him,  and  the  sound  of  his  gongs  came  (over  the  lake).  She  asked, 
"Whom  have  they  speared?"  and  (the  people)  said,  "They  have 
speared  your  husband."^ 

^  The  confusion  which  occurs  here  between  the  two  women  is  typical 
of  texts  dictated  by  the  latmul.    It  is  evident  that  my  informant  was 


146      The  Ethos  of  latmul  Culture:  the  Women 

Then  she  filled  up  a  string  bag  with  shell  valuables  and  she  (went 
to  the  ceremonial  house  and)  said,  "Men  of  this  village,  I  have 
brought  (valuables)  for  you."  But  they  said,  "  No.  We  do  not  want 
them",  and  they  were  ashamed  (because  they  had  not  dared  to 
accept  the  valuables  which  she  had  offered  as  payment  for  assist- 
ance). 

Then  she  went  down  into  her  canoe;  she  loaded  the  valuables 
into  the  canoe;  she  took  off  her  skirt  and  put  it  in  the  canoe. 
Au-vitkai-mangka  was  in  the  stern  and  Tshanggi-mbo  in  the  bows. 
The  bag  of  valuables  was  in  the  middle  of  the  canoe.  She  went 
up  the  river  to  Palimbai,  because  she  had  heard  his  gongs.  The  two 
of  them  went  by  night. 

They  sat  leaning  against  the  ceremonial  mound  (a  place  of  refuge) 
in  Palimbai,  and  they  put  the  bag  of  valuables  on  the  ground  close 
to  the  mound.  At  dawn  (the  people  of  Palimbai)  got  up  and  saw 
(them).  They  were  sitting  stripped  of  their  skirts,  with  their  skirts 
on  their  shoulders.^ 

♦The  men  of  Palimbai  said,  "They  are  women  of  Kararau"; 
and  they  were  for  spearing  them.  The  women  said,  "Why  will 
you  spear  us? ".  Kaulievi  (of  Palimbai)  saw  and  said,  "  Don't  spear 
them" ;  and  he  said  "  Come".  Then  he  beat  the  gong  to  summon 
all  the  men  of  Palimbai,  Kankanamun,  Malinggai,  and  Jentschan. 
The  men  of  the  four  villages  came  together,  and  the  women  told 
them  to  debate.  The  men  said,  "What  women  are  you?";  and  the 
women  said,  "We  are  women  of  lenmali."  (lenmali  is  the  name 
of  the  old  site  of  Mindimbit.) 

Kaulievi  said,  "Tell  your  story";  and  Au-vitkai-mangka  said, 
"That  is  (the  sound  of)  my  brother's  gongs  coming" ;  and  the  men 
said,  "Who  speared  him?". 

Au-vitkai-mangka  then  (calling  the  names  of  the  totems  of  the 
four  villages)  appealed  to  Kankanamun:  "You!  Crocodile!  Wani- 
mali!";  and  to  Malingai:  "You!  Crocodile!  Kavok!";  and  to 
Palimbai:  "You!  Pig!  Palimbai-awan !  " ;  and  to  Jentschan:  "You! 
Pig!  Djimbut-nggowi!".  And  she  said,  "I  shall  take  away  my  bag 
of  valuables." 

She  set  out  the  valuables  in  a  line ;  and  the  four  villages  accepted 
them.  That  night  they  debated,  "Already  tomorrow  we  shall  raid 
them."  Each  of  the  four  villages  (brought)  a  fleet  of  canoes.  They 
formed  into  one  fleet  on  the  Sepik  River.* - 

They  (the  men)  gave  a  spear  to  Au-vitk*ai-mangka  and  the  men  of 

describing  the  man's  sister  as  hearing  the  sound  of  the  gongs  while  she 
was  fishing,  but  when  she  asks  who  has  been  killed  the  reply  is  "  Your 
husband". 

*  The  nakedness  of  the  women  in  this  context  seemed  so  natural  to 
me  when  I  was  told  the  myth,  that  I  did  not  enquire  into  the  reasons 
for  it.  I  have  no  doubt,  however,  that  this  nakedness  is  the  mark  of  the 
suppliant  and  that  it  is,  in  some  degree,  analogous  to  the  nakedness  of 
the  women  in  naven  when  they  lie  down  before  the  hero. 

^  The  passage  between  asterisks  has  been  condensed. 


The  Ethos  of  latmul  Culture:  the  Women      147 

Palimbai  gave  another  spear  to  Tshanggi-mbo.  They  gave  one  canoe 
to  the  two,  a  swift  canoe;  and  the  two  women  were  in  the  centre 
of  the  fleet. 

When  they  drifted  down  to  the  Kararau  (reaches  of  the) 
Sepik,  (the  canoes  took  up  formation*).  The  two  women  hid  in 
the  centre.  Then  the  men  shot  an  eel.  It  said  "War".  (A  favour- 
able omen;  and  here  my  informant  reproduced  the  grunting  of 
the  eel.) 

The  two  women  came  out  (of  the  fleet).  They  (went  forward 
and)  sang  dirges  in  midstream.^  They  were  smeared  with  clay  (for 
mourning)  and  the  people  of  Kararau  came  out  to  spear  them.  But 
(the  women)  were  going  down  stream  in  a  swift  canoe.  They  went 
for  the  setting  of  Palimbai's  battle  (i.e.  the  women  acted  as  decoys). 
The  Palimbai  people  killed  the  people  of  Kararau  and  they  caught 
two  men  (alive)  in  their  hands.  Au-vitkai-mangka  speared  one  of 
them.  Tshanggi-mbo  speared  the  other.  They  speared  them  all, 
every  one  of  them. 

They  all  went  upstream  together  to  Palimbai  and  there  they  beat 
the  gongs.  The  two  women  beat  the  gongs. ^  In  the  morning  the 
men  cut  the  women's  hair  and  oiled  them  and  presented  valuables 
to  them.  Then  they  brought  the  women  in  a  fleet  of  canoes  to 
lenmali,  and  left  them.  Kaulievi  said  "Kararau  are  our  enemies" 
and  he  came  and  set  up  a  stone  (in  lenmali,  now  removed  to 
Mindimbit).  And  so  Palimbai  took  Kararau  for  enemies.  That  is 
the  stone,  here,  and  the  name  of  the  stone  is  Kaulievi,  an  ancestor 
of  Kepmaindsha.  That  is  why  Kepmaindsha  came  here,  and 
Tonggalus  too.  (Kepmaindsha  and  Tonggalus  were  two  men  who 
had  left  their  own  villages  as  a  result  of  quarrels.)  Later  when  the 
two  women  died,  they  made  a  song  about  them ;  the  clan  Mwailambu 
(to  which  Au-vitkai-mangka  and  my  informant  belonged)  made  the 
song. 

In  the  household,  too,  a  woman  may  have  considerable 
power  and  authority.  She  it  is  who  feeds  the  pigs  and  catches 
the  fish;  and  it  is  upon  these  activities  that  her  husband 
chiefly  depends  for  the  wealth  which  helps  him  to  make  a 
splash  in  the  ceremonial  house.  When  a  man  is  haggling 
silently  over  a  canoe  or  sleeping  bag,  he  will  withdraw 
before  concluding  the  deal,  in  order  to  consult  his  wife.  And, 
judging  by  the  things  which  the  wife  is  reported  to  say  in 

*  This  phrase  is  substituted  for  a  list  of  technical  terms  which  do  not 
concern  us  here. 

*  It  is  usual  for  widows  to  sing  dirges  for  their  husbands  as  they  go 
in  their  canoes  to  and  from  their  work  (cf.  p.  157). 

*  The  beating  of  the  gongs,  like  the  business  of  head-huntipg,  is 
normally  only  done  by  men.  The  oiling,  hair-cutting  and  presentation  of 
valuables  are  expressions  of  respect. 


148     The  Ethos  of  latmul  Culture:  the  Women 

such  circumstances,  it  appears  that  wives  hold  the  purse 
strings  very  tight.  But  the  stubbornness  of  an  absent  wife 
makes  a  very  convenient  tool  in  the  business  of  striking  a 
bargain  and  I  doubt  whether  the  wives  are  really  as  "  strong  " 
as  their  husbands  report. 

In  a  few  households,  however,  it  is  definitely  the  wife  who 
"wears  the  trousers",  and  in  two  such  cases  the  sympathy 
of  outsiders  went  to  the  wife  rather  than  to  the  henpecked 
husband.  It  was  the  wife's  misfortune  to  have  married  a 
weakling. 

But,  as  against  the  occasional  instances  in  which  women 
take  up  an  assertive  role  and  even  participate  in  warfare,  the 
more  habitual  emphasis  of  the  women's  ethos  is  upon  quiet 
co-operative  attitudes.  Though  the  woman  may  take  the 
initiative  in  sexual  advances,  it  is  the  activity  of  the  male 
which  is  stressed  in  the  native  remarks  about  copulation, 
while  the  part  played  by  the  female  is  despised.  In  the 
latmul  language  the  ordinary  verb  for  copulation  and  the 
jocular  synonyms  which  are  used  for  it  are,  so  far  as  I  know, 
all  of  them  transitive  and  in  their  active  forms  refer  to  the 
behaviour  of  the  male.  The  same  verbs  may  be  used  of  the 
female  role,  but  always  in  the  passive.^ 

Thus,  in  our  study  of  the  women's  ethos,  we  find  a  double 
emphasis.  For  the  most  part,  the  women  exhibit  a  system  of 
emotional  attitudes  which  contrasts  sharply  with  that  of  the 
men.  While  the  latter  behave  almost  consistently  as  though 
life  were  a  splendid  theatrical  performance — almost  a  melo- 
drama— with  themselves  in  the  centre  of  the  stage,  the  women 
behave  most  of  the  time  as  though  life  were  a  cheerful  co- 
operative routine  in  which  the  occupations  of  food-getting 
and  child-rearing  are  enlivened  by  the  dramatic  and  exciting 
activities  of  the  men.  But  this  jolly,  co-operative  attitude  is 
not  consistently  adopted  in  all  contexts,  and  we  have  seen 
that  women  occasionally  adopt  something  approaching  the 
male  ethos  and  that  they  are  admired  for  so  doing. 

In  the  ceremonial  activities  of  the  women,  the  same  double 

^  These  verbs  may  also  be  used  in  the  dual  in  the  active  voice.  The 
pidgin  English  idom,  "play",  has  not  been  adopted  in  literal  translation 
into  latmul. 


The  Ethos  of  latmul  Culture:  the  Women      149 

emphasis  is  present,  and  these  activities  fall  into  two  distinct 
ethological  groups  according  as  one  or  the  other  emphasis 
is  predominant.  In  general,  the  jolly,  co-operative  emphasis 
is  most  evident  when  women  celebrate  by  themselves  in  the 
absence  of  men,  while  the  proud  ethos  is  exhibited  when 
women  celebrate  publicly  in  the  dancing  ground  of  the  village 
with  men  in  the  audience. 

In  the  first  group  are  the  frequent  dances  held  by  women 
in  the  houses.  These  dances  are  very  much  resented  by  the 
men,  who  regard  them  with  contempt  and  do  all  they  can  to 
discourage  them.  When  the  women's  ceremonial  demands 
that  the  performers  shall  keep  taboos  on  sexual  intercourse, 
the  men  do  their  best  to  cause  them  to  break  these  taboos — 
and  then  boast  to  the  anthropologist  of  the  postponement 
of  the  ceremony.  At  such  times  the  sex  opposition — never 
far  from  the  surface — comes  to  a  head.  Quarrels  between 
husbands  and  wives  are  especially  frequent  and  the  wives 
take  their  revenge  by  refusing  to  cook  sago  for  their  hus- 
bands. 

I  found  a  husband  sitting  sulkily  in  the  ceremonial  house. 
He  had  a  lump  of  sago  which  he  was  roasting  rather  in- 
effectually, naked  on  the  fire — for  the  men  believe  that  the 
art  of  cooking  pancakes  of  sago  is  one  which  they  cannot 
learn.  He  said,  "Yes,  we  copulate  with  them,  but  they  never 
retaliate",  a  reference  to  the  despised  passive  sexual  role. 
Then  he  jumped  to  his  feet  and  shouted  this  taunt  across 
the  village  to  the  women  in  his  house,  from  which  he  was 
excluded. 

For,  in  spite  of  their  contemptuous  attitude,  the  men 
withdraw  quietly  enough  when  the  dance  is  actually  about 
to  take  place ;  and  the  women  are  left  in  complete  command 
of  the  house.  They  remove  all  the  sleeping  bags,  clearing  the 
floor-space  for  the  dance ;  and  a  great  crowd  of  women  col- 
lects together  from  all  over  the  village,  all  joking  and  in  the 
best  of  humour.  After  a  while  the  dancing  begins  and  the 
gathering  sounds  very  gay. 

On  one  occasion  I  was  sitting  in  the  ceremonial  house  when 
these  sounds  reached  us.  The  men  greeted  the  sound  with 
contempt ;  but  I  asked  if  I  could  go  and  look  at  the  women's 


150      The  Ethos  of  latmul  Culture:  the  Women 

dance.  The  men  told  me  that  the  dances  were  very  silly  and 
not  worth  my  looking  at,  that  they  could  not  compare  with 
the  dances  of  the  men.  I  said  that  Mindimbit  was  a  poor, 
"  cold  "  village ;  the  men  never  performed  any  of  their  vaunted 
spectacles;  and  if  I  couldn't  see  any  dancing  I  was  going  to 
go  to  another  village.  Finally  and  very  reluctantly  a  young 
man  said  he  would  take  me  to  the  dance  and  we  went  together. 
We  entered  while  the  women  were  dancing  round  the  floor 
in  short  jumping  steps  and  singing  a  rather  catchy  quick 
tune.  We  sat  down  on  stools  in  a  corner  of  the  floor  in  silence. 
My  companion  was  acutely  uncomfortable  and  after  a  few 
minutes  he  slipped  away. 

When  the  women  had  finished  the  song  they  came  crowding 
up  to  me  and  off^ered  me  areca  nut  and  betel.  I  asked  for 
lime.  Most  of  them  had  never  seen  me  chew  betel  and  my 
acceptance  of  the  offer  created  some  excitement,  screams  of 
laughter  and  noisy  screaming  talk — like  a  flock  of  parrots. 
In  the  middle  of  this  excitement  two  women  started  to  dance 
in  front  of  me.  They  stood  face  to  face  in  a  jumping  dance 
and  at  every  jump  one  woman  pushed  her  hands  forward 
with  palms  pressed  together,  the  other  woman  received  the 
hands  between  her  own.  Between  the  beats  the  first  woman 
drew  her  hands  back  only  to  push  them  forward  again  at  the 
next  beat.  This  dance  was  obviously  a  representation  of 
copulatory  action,  but  I  was  completely  surprised  when  sud- 
denly, in  a  single  jump,  the  two  women  dropped  to  a  sitting 
position  on  the  floor,  still  facing  each  other  and  one  sitting 
between  the  legs  of  the  other,  one  of  the  standard  positions 
for  sexual  intercourse.  In  this  position  the  women  went 
through  the  motions  of  copulation  still  in  time  to  the  beat 
of  the  song;  and  then,  as  suddenly  as  they  had  sat  down, 
they  jumped  up  in  a  single  motion  and  after  a  few  more 
jumps  they  broke  off,  giggling. 

The  mere  description  of  what  these  two  women  did  gives 
very  little  idea  of  the  extraordinary  naivete  of  this  "  obscenity  " 
and  the  contrast  between  it  and  the  harsher  obscenity  of  the 
men.  Lacking  a  photographic  record  I  can  only  record  my 
subjective  impression  of  this. 

After  more  joking  the  crowd  of  women  left  me  to  continue 


The  Ethos  of  latmul  Culture:  the  Women      151 

their  dancing  round  the  house.  The  same  jolly  atmosphere 
continued  and  I  had  no  doubt  that  this  was  the  regular  tone 
of  the  women's  dances  in  the  absence  of  men.  I  think  too 
that  an  analysis  of  the  tunes,  sung  by  women  and  men 
respectively,  would  show  the  same  ethological  contrast  be- 
tween the  sexes  which  I  observed  in  their  behaviour. 

But  this  characteristic  jollity  is  not  carried  over  to  those 
occasions  when  women  celebrate  publicly.  Then  they  march 
in  procession  in  the  middle  of  the  dancing  ground  before  a 
mixed  audience  of  men  and  other  women,  they  are  fully 
decorated  and  wear  among  other  things  many  ornaments 
usually  worn  by  men — a  sort  of  mild  transvesticism  which 
will  be  referred  to  again  in  the  theoretical  analysis  of  naven. 
They  march  with  a  fine  proud  bearing  very  different  from 
their  jolly  behaviour  when  men  are  absent,  and  different  too 
from  their  quieter  demeanour  on  everyday  occasions  when  they 
are  in  the  presence  of  men  but  are  not  decked  out  in  finery. 
Their  marching  gait  in  these  processions  is  indeed  more 
closely  comparable  with  their  swaggering  demeanour  when 
dressed  in  full  homicidal  war  paint  for  the  naven  ceremonies 
than  with  their  patterns  of  behaviour  on  other  occasions. 

Examples  of  this  proud  bearing  and  of  the  mild  trans- 
vesticism are  shown  on  Plate  XIX,  and  a  similar  phenomenon 
to  that  which  appears  on  ritual  occasions  may  be  observed 
fairly  constantly  under  the  experimental  conditions  produced 
by  pointing  a  camera  at  an  individual.  When  a  woman  is 
photographed,  her  response  to  the  camera  depends  on 
whether  she  is  wearing  her  finery  or  is  in  everyday  dress. 
In  her  finery  she  holds  her  head  high  when  the  eye  of  the 
camera  is  on  her;  but  in  everyday  dress  she  hangs  her  head 
and  rather  shrinks  from  the  public  appearance  constituted  by 
standing  up  alone  before  the  photographer  while  her  friends 
are  watching  in  the  background  (cf.  Plate  XXVI).  When 
a  man  is  photographed,  whatever  his  costume,  he  tends  to 
swagger  before  the  camera  and  his  hand  goes  almost  instinc- 
tively to  his  lime  stick  as  if  about  to  make  with  it  the  loud 
grating  sound  which  is  used  to  express  anger  and  pride. 


CHAPTER  XI 

Attitudes  towards  Death 

WE  have  so  far  examined  the  behaviour  of  men  and 
women  only  in  the  everyday  and  the  ceremonial 
contexts  of  their  culture.  But  the  contrast  between 
the  sexes  is  even  more  striking  when  the  individuals  are  faced 
with  events  highly  charged  with  emotion.  To  illustrate  this 
I  shall  describe  the  sequences  of  events  after  a  death  has 
occurred. 

In  Palimbai  I  was  woken  up  at  about  4.30  a.m.  one  morning 
by  the  sound  of  weeping  in  the  house  next  to  mine.  I  went 
to  see  what  was  happening  and  found  that  a  young  man,  who 
had  been  sick  for  two  or  three  months,  had  finally  died.  The 
corpse  was  stretched  out  straight  and  was  naked.  A  circle 
of  women  were  crouched  around  it,  and  the  mother  of  the 
dead  man  had  the  head  in  her  lap.  A  fire  was  burning  close 
to  the  corpse,  and  gave  the  only  light  in  the  house.  The 
women  were  quietly  weeping  and  dirging,  singing  songs  of 
the  dead  man's  maternal  clan — songs  which  might  be  used 
on  gay  and  everyday  occasions,  only  now  the  singing  was 
slow  and  out  of  tune  and  broken  with  sobs. 

From  time  to  time  there  were  pauses  when  all  were  quiet, 
and  then  one  of  the  women  would  make  some  remark  about 
the  dead  man.  Some  incident  of  his  life  was  referred  to, 
or  some  small  possession  of  his  mentioned  with  the  suggestion 
that  it  should  be  buried  with  him.  Then  the  songs  and  the 
sobs  were  resumed,  set  off  by  this  recall  of  another  facet  of 
their  personal  loss. 

There  was  one  man  in  the  house.  He  was  sitting  apart 
from  the  group  of  women,  silent  and  embarrassed.  When  I 
went  up  and  spoke  to  him  he  greeted  my  intrusion  with 
pleasure  and  was  very  ready  to  discuss  the  arrangements  for 
the  funeral — how  the  Government  had  forbidden  them  to 
expose  the  corpse  in  a  canoe  till  the  floods  abated ;  as  it  was, 
they  would  have  to  take  the  body  into  the  Tshuosh  country 


Attitudes  towards  Death  153 

to  find  dry  ground  for  the  burial;  they  would  measure  the 
body  in  order  to  know  how  big  a  hole  to  make ;  and  so  on. 

The  women's  weeping  continued  until  after  dawn,  but  it 
was  no  hysterical  exaggeration  of  grief  such  as  is  recorded 
from  other  primitive  communities.  My  feeling  was  that  I  was 
witnessing  an  easy  and  natural  expression  of  sorrow  at  a 
personal  loss. 

The  behaviour  of  the  man  was  in  marked  contrast  to  this. 
On  the  one  hand,  he  quite  evidently  wanted  to  escape  from 
his  embarrassment  into  conversation  about  the  funeral,  and, 
on  the  other  hand,  he  boasted:  "We  (latmul)  are  not  people 
who  only  play  at  weeping",  and  when  I  asked  whether  it  was 
not  only  the  women  who  wept  he  felt  this  remark  as  an 
aspersion  on  the  men,  and  insisted  that  men  also  weep.  Later 
in  the  conversation  he  turned  his  attention  from  the  affairs 
cormected  with  the  death  and  started  to  lecture  me  on  the 
East  wind  and  its  totemic  position. 

After  dawn,  we  first  waited  for  the  rain  to  stop  and  then 
proceeded  with  the  funeral,  for  the  men  say:  "Tears  are  not 
found  in  the  lake",  meaning  that  the  supply  will  not  last 
long  and  therefore  they  must  bury  the  body  quickly.  The 
men  put  the  body  in  a  canoe  and  took  it  over  the  fens 
from  one  piece  of  supposedly  higher  ground  to  another, 
but  all  were  flooded.  We  were  a  party  of  eight  of  whom  two 
were  women,  the  mother  and  sister  of  the  dead  man.  The 
mother  sat  immediately  behind  the  corpse,  sometimes  dirging 
over  it.  Conversation  on  the  journey  was  quiet  and  concerned 
with  possible  causes  of  the  death.  Our  plan  was  to  take  the 
body  to  Marap  village,  but  the  men  were  impatient  and  on 
the  way  said:  "No,  Marap  is  a  long  way.  Let  us  bury  him 
on  Movat  Tevwi "  (a  piece  of  high  ground  in  the  fens) :  but 
in  the  end  we  had  to  go  to  Marap  and  arrived  there  in  the  late 
afternoon.  The  Tshuosh  were  not  pleased  to  see  us,  but  at 
last  they  permitted  the  body  to  be  buried  under  a  deserted 
house. 

The  men  had  some  difficulty  in  digging  a  grave  and  the 
site  had  to  be  changed  twice  because  they  came  on  other  old 
bones  when  they  dug.  Finally  the  corpse  was  laid  in  the 
grave  and  the  portrait  skull  of  the  dead  man's  brother  was 


1^4  Attitudes  towards  Death 

deposited  with  him.  A  shilling  was  placed  in  each  of  his 
hands,  and  his  string  bag  was  placed  in  the  grave.  Since 
he  was  buried  among  the  Tshuosh,  the  grave  was  oriented 
so  that  (on  raising  his  head)  the  corpse  would  look  towards 
the  setting  sun  as  is  the  custom  of  the  Tshuosh.  In  Palimbai 
the  dead  are  normally  buried  with  their  feet  towards  the 
dancing  ground,  so  that  the  corpse  is  not  "looking  into  the 
bush". 

The  women  retained  their  skirts  throughout  this  burial, 
but  I  was  told  by  an  informant  in  Kankanamun  that  the 
mother,  sister  and  wife  of  a  dead  man  would  normally  be 
naked  while  he  was  being  buried.  This  nakedness  is,  no 
doubt,  in  some  way  analogous  to  the  nakedness  of  the  women 
when  they  lie  down  before  the  hero  in  naven,  and  with  the 
nakedness  of  female  suppliants. 

The  mother  of  the  dead  man  stayed  behind  in  Marap  to 
mourn  for  a  few  days,  but  the  rest  of  us  returned  that 
evening  to  Palimbai  as  a  normally  cheerful  party,  no  longer 
an  embarrassed  and  silent  group. 

A  second  occasion  on  which  I  was  able  to  observe  the 
reactions  of  the  men  to  a  death  was  on  the  day  after  Tep- 
managwan,  a  great  fighter  of  Palimbai,  had  died.  He  died 
during  the  night  and  was  buried  in  the  early  morning. 
I  arrived  in  the  village  at  about  9  o'clock,  after  the  inter- 
ment, and  found  that  the  men  had  by  then  left  the  grave 
and  gone  to  the  ceremonial  house.  A  few  women  were 
weeping  at  the  graveside,  and  from  the  ceremonial  house 
I  could  just  hear  the  weeping  of  others  in  the  house  of  the 
dead  man. 

I  proposed  going  to  the  house,  but  the  men  hinted  that 
I  should  not.  They  were  just  starting  a  debate  in  the  cere- 
monial house.  It  was  a  scandal  that  Tepmanagwan  had  died 
without  passing  on  his  esoteric  knowledge,  and  the  debate 
was  an  enquiry  as  to  whose  fault  this  was.  A  few  men  sobbed 
while  they  were  making  their  speeches,  and  I  found  it  hard 
to  judge  whether  these  sobs  were  the  result  of  genuine  feeling 
bursting  its  way  to  the  surface  against  resistance,  or  whether 
they  were  a  theatrical  performance  staged  in  absence  of  strong 
feeling  to  give  this  impression.   In  any  case,  it  was  perfectly 


Attitudes  towards  Death  155 

clear  that  the  men's  sobs  were  very  far  removed  from  the 
natural  weeping  of  the  women. 

The  debate  reached  no  conclusion  and,  when  it  petered 
out,  the  men  set  up  a  figure  to  represent  the  dead  man  (cf. 
Plate  XX  a).  The  head  of  the  figure  was  an  unripe  coconut 
and  the  body  was  made  of  bundles  of  palm  leaves.  Spears 
were  set  up  against  the  figure  with  their  points  stuck  into 
it  to  mark  where  the  man  had  been  wounded  in  war,  and 
other  spears  were  stuck  into  the  ground  beside  the  figure 
for  those  which  he  had  dodged.  A  series  of  vertical  spears 
was  set  up  in  front  of  the  figure  according  to  his  achieve- 
ments. The  figure  itself  was  ornamented  with  shells,  etc. 
Six  sago  baskets  were  suspended  from  its  right  shoulder  to 
represent  his  six  wives,  a  string  bag  was  suspended  on  the 
left  shoulder  representing  his  skill  in  magic.  A  number  of 
sprigs  of  ginger  in  its  headdress  represented  persons  whom 
he  had  invited  to  the  village  so  that  other  people  could  kill 
them.  In  the  right  hand  of  the  figure  was  a  dry  lump  of  sago, 
because  it  was  said  that  in  his  lifetime  he  had  once  killed  a 
bird  by  throwing  a  lump  of  sago  at  it.  A  branch  of  timbut 
(lemon)  set  in  the  ground  beside  the  figure  was  symbolic  of 
his  knowledge  of  mythology.  Finally,  on  the  ground  at  the 
feet  of  the  figure  were  a  broom  and  a  pair  of  boards  used  for 
picking  up  rubbish.  These  objects  were  symbolic  of  the  work 
which  the  dead  man  had  done  in  cleaning  the  ceremonial 
house  during  his  lifetime. 

This  figure  was  set  up  by  members  of  the  initiatory  moiety 
of  which  the  deceased  was  a  member.  It  was  a  boast  of  the 
greatness  of  their  moiety,  and  when  the  figure  was  completed 
all  the  men  of  both  moieties  crowded  round  it.  The  members 
of  the  opposite  moiety  came  forward  one  by  one  to  claim 
equivalent  feats.  One  man  said:  "I  have  a  wound  here  on 
my  hip,  where  the  (people  of)  Kararau  speared  me.  I  take 
that  spear",  and  took  the  spear  set  against  the  figure's  hip. 
Another  said:  "I  killed  so-and-so.  I  take  that  spear",  and 
so  on  till  all  the  emblems  of  prowess  had  been  removed. 

Thus  the  men  made,  out  of  the  context  of  death  only  a 
few  hours  old,  an  occasion  for  expressing  the  competitive 
pride  of  the  initiatory  moieties.  They  escaped  entirely  from 


156  Attitudes  towards  Death 

a  situation  which  was  embarrassing  because  it  seemed  to 
demand  a  sincere  expression  of  personal  loss,  an  expression 
which  their  pride  could  scarcely  brook.  From  this  situation 
they  took  refuge  in  a  cultural  stunt.  They  re-phrased  their 
attitude  towards  the  dead  and  expressed  it  satisfactorily  in 
terms  of  spectacular  pride,  the  emotional  language  in  which 
they  are  at  ease.^  But  such  a  handling  of  grief  is,  I  think, 
still  not  adequate,  and  later  a  further  compensation  is  added. 
It  is  my  impression^  that  when  a  man  is  asked  about  some 
past  funeral  he  will  generally  drag  into  his  answer  some 
reference  to  his  own  great  weeping,  in  spite  of  the  fact  that, 
at  the  time,  he  wept  but  little  and  probably  made  a  show  of 
his  resistance  to  womanly  tears. 

In  the  later  mortuary  ceremonies  the  contrast  between  the 
behaviour  of  the  two  sexes  continues.  The  skull  of  the  dead 
man  is  exhumed  and  a  portrait  is  modelled  upon  it  in  clay. 
This  is  set  up  one  night  as  the  head  of  a  highly  decorated 
doll  which  represents  the  deceased  (cf.  Plate  XXI  b).  Around 
this  figure  the  men  stage  an  elaborate  performance  of  name 
songs  and  flute  music.  The  ceremony  which  is  called  min- 
tshanggu  (cf.  p.  47)  takes  place  in  a  dwelling  house  and  the 
women  are  present  as  audience.  The  flutes  are  played  by 
men  hidden  under  the  platform  on  which  the  figure  stands, 
whither  they  have  been  secretly  smuggled.  Thus  though  the 
ceremony  takes  place  in  a  dwelling  house  and  its  context  is  a 
personal  one,  it  is  staged  upon  the  same  general  principles 
as  all  the  other  performances  staged  by  the  men,  a  spectacle 
for  the  admiration  and  mystification  of  the  women. 

Later  the  women  have  a  little  mourning  ceremony  by 
themselves  in  the  absence  of  men.  This  is  called  yigen  kundi 
("  quiet  singing").  It  takes  place  at  night  in  a  house  from  which 
the  men  have  withdrawn.  A  little  food  is  hung  up  for  the 
ghost  to  "eat"  and  the  women  sit  in  a  circle  by  the  firelight 

^  In  our  own  culture,  of  course,  both  these  types  of  emotional  pattern 
and  many  others  are  mixed  and  tangled  together  in  our  mortuary  cere- 
monial. Culture  contacts  and  the  recurring  instability  of  Western  Euro- 
pean societies  have  provided  us  with  every  sort  of  conflicting  phrasing, 
and  these  phrasings  have  been  preserved  for  us  in  script  through  the 
ages.    But  the  latmul  have  a  less  confused  culture. 

^  Unfortunately  I  took  no  notes  of  such  effusions,  accepting  them  as 
a  matter  of  course. 


Attitudes  towards  Death  157 

and  softly  sing  the  name  songs  of  the  dead  man's  mother's 
clan.  The  wife  or  mother  of  the  dead  may  weep  a  little,  but 
the  general  tone  of  the  group  is  one  of  quiet  sorrow  rather 
than  of  passionate  grief.  The  "quiet  singing"  goes  on  till 
late  in  the  night,  when  the  women  disperse  to  their  houses. 

In  the  months  or  years  which  follow,  the  mother  or  wife 
of  the  dead  man  will  occasionally,  when  she  is  alone,  sing 
as  a  dirge  one  of  the  name  songs  of  his  maternal  clan ;  these 
dirges  may  often  be  heard  on  the  river,  coming  from  some 
woman  mourning  as  she  paddles  her  canoe  to  her  garden  or 
fish  traps.  The  men  quite  frequently  caricature  this  musical 
effort,  probably  because  the  attitude  of  the  women  towards 
death  is  one  which  they  themselves  find  distasteful. 

Indeed  one  of  the  most  important  phenomena  which  is 
brought  to  light  by  examination  of  ethological  contrast  is  this 
distaste  which  persons  trained  in  one  ethos,  their  emotional 
reactions  standardised  in  one  pattern,  feel  for  other  possible 
ethoses.^  In  the  illustration  which  I  gave,  I  mentioned  how 
a  remark  which  is  out  of  tune  with  the  temporary  ethos  of  a 
group  of  Englishmen  is  received  with  silence,  and  in  latmul 
culture  we  may  recognise  the  same  phenomenon  in  the 
distaste  which  the  men  feel  for  the  ethos  of  the  women.  This 
phenomenon  is  extraordinarily  widespread  and  it  affects  even 
the  anthropologist  whose  task  it  is  to  be  an  impartial  student 
of  ethos.  Every  adjective  which  he  uses  is  coloured  by  and 
evokes  the  feelings  which  one  sort  of  personality  has  about 
another.  I  have  described  the  ethos  of  the  men  as  histrionic, 
dramatising,  over-compensating,  etc.,  but  these  words  are 
only  a  description  of  the  men's  behaviour  as  seen  by  me, 
with  my  personality  moulded  to  a  European  pattern.  My 
comments  are  in  no  sense  absolute  statements.  The  men 
themselves  would  no  doubt  describe  their  own  behaviour  as 
"natural";  while  they  would  probably  describe  that  of  the 
woman  as  "sentimental". 

It  is  difficult  too  to  describe  a  pair  of  contrasting  ethoses 

^  The  same  phenomenon  is  to  be  observed  also  in  the  reactions  of 
individuals  trained  in  one  system  of  cultural  structure  to  other  possible 
systems.  They  are,  I  think,  liable  to  regard  the  other  systems  as  non- 
sensical, illogical  and  perhaps  tedious. 


158  Attitudes  towards  Death 

without  so  weighting  the  descriptions  that  one  or  the  other 
appears  preferable  or  more  "natural".  The  business  of  the 
scientist  is  to  describe  relationships  between  phenomena,  and 
any  ethos  which  he  finds  in  a  culture  must  be  regarded  not 
as  "natural"  but  as  normal  to  the  culture.  Unfortunately 
what  is  normal  to  one  culture  may  well  be  abnormal  to 
another,  and  the  anthropologist  has  at  his  disposal  only  the 
adjectives  and  phrases  of  his  own  culture.  Thus  it  has  hap- 
pened that  English  people  with  whom  I  have  discussed 
latmul  ethos  have  sometimes  remarked  that  the  women 
appear  to  be  "well  adjusted"  while  the  men  appear  to  be 
"  strained  "  and  "  psychopathic  ".  My  friends  forget  that  the 
values  assigned  by  European  psychiatrists  to  various  mental 
conditions  are  either  cultural  values  based  upon  European 
ethos  or  estimates  of  the  fitness  of  the  individual  for  life  in 
a  European  community. 

The  pride  of  the  men,  when  seen  in  contrast  with  the 
women's  ethos,  may  appear  to  my  readers  somewhat  angular 
and  uncomfortable.  I  found  it  also  splendid.  I  have  not 
stressed  this  aspect  enough,  and  therefore  I  shall  conclude 
the  "description  of  latmul  ethos  with  a  free  translation  of  a 
story  which  illustrates  how  a  man  should  behave  when  his 
own  death  stares  him  in  the  face : 

A  man  went  with  his  dog  to  hunt  for  wild  pigs  in  the  sago 
swamps.  When  they  had  killed  a  pig  the  man  went  to  wash 
its  guts  in  a  lake.  While  he  was  doing  this  a  giant  crocodile 
(Mandangku,  an  ancestor  of  Tshingkawi  clan)  seized  him  by 
the  instep  and  held  him  fast. 

The  man  said  to  the  dog,  "Go  home  and  sniff  at  my 
feather  headdress,  and  sniff  at  my  armbands  and  all  my 
ornaments." 

Then  the  dog  went  home  and  when  the  man's  wife  saw 
the  dog  sniffing  at  the  ornaments  she  took  them  and  put  them 
in  a  basket,  and  the  dog  led  her  and  her  child  back  to  where 
the  man  was,  still  held  fast  by  the  crocodile  on  the  lake  side. 

When  he  saw  them  the  man  said,  "My  child,  my  wife, 
I  am  lost",  and  then  he  said,  "  Give  me  my  things." 

He  put  on  his  legbands  and  his  shell  girdle.  He  put  on 
one  of  his  armbands  and  then  he  put  on  the  other.   He  hung 


Attitudes  towards  Death  159 

his  mother-of-pearl  crescent  round  his  neck.  Finally  he  put 
on  his  headdress  of  parrot  skins  and  bird-of-paradise  feathers. 
Then  he  said  to  his  wife  and  child,  "  Come  close  and  wait." 

The  crocodile  began  to  pull  him  down  into  the  water.  He 
took  off  his  legbands  and  threw  them  ashore.  The  crocodile 
pulled  him  farther  and  he  took  off  his  shell  girdle.  The 
crocodile  pulled  him  farther  till  the  water  came  level  with 
his  armpits,  and  then  he  took  off  his  armbands  and  threw 
them  ashore.  The  crocodile  pulled  him  farther,  and  finally 
he  took  off  his  mother-of-pearl  crescent  and  his  feather  head- 
dress and  threw  them  ashore.  He  said,  "It  is  done",  and 
then  he  said,  "Go!  my  wife,  my  child,  go!  What  is  become 
of  me.?" 

Then  there  was  the  sound  of  splashing,  the  crocodile 
waved  its  tail  and  bits  of  leaf  and  grass  were  stirred  up  from 
the  bottom  of  the  water. 


CHAPTER  XII 

The  Preferred  Types 

SHORTLY  after  my  return  from  New  Guinea  I  read  for 
the  first  time  Kretschmer's  Physique  and  Character 
(English  translation,  1925),  and  it  was  at  once  obvious 
that  the  contrast  which  I  had  observed  between  the  sexes  in 
latmul  culture  was  in  some  ways  comparable  with  Kretsch- 
mer's contrast  between  cyclothyme  and  one  group  of  schizo- 
thyme  personalities.^  It  appeared  that  schizothyme  behaviour 
was  "fashionable",  standardised  in  latmul  culture  as  appro- 
priate to  men ;  while  cyclothyme  behaviour  was  standardised 
for  women.  The  equivalence  between  the  schizothyme  be- 
haviour of  the  latmul  men  and  that  of  certain  natives  of 
South  Germany  is  only  partial ;  and  the  exact  analysis  of  the 
relationship  between  latmul  schizothymia  and  German  schizo- 
thymia  requires  a  great  deal  more  research  and  a  knowledge 
of  the  possible  standardisations  of  schizothyme  ethos  in  other 
cultures. 

In  the  case  of  the  equivalence  between  the  latmul  women's 

^  I  should  perhaps  state  my  opinion  of  Kretschmer's  dichotomy.  I  be- 
lieve that,  though  the  external  world  is  perhaps  not  built  upon  a  dualistic 
basis,  dualisms  and  dichotomies  provide  a  convenient  technique  for 
describing  it,  and  this  technique  is  so  standardised  in  our  culture  that 
there  is  little  hope  of  avoiding  it.  I  agree,  of  course,  with  Kretschmer 
that  we  must  not  think  of  these  dichotomies  as  discontinuous,  but  rather 
that  we  should  expect  to  classify  individuals  on  a  scale  varying  between 
the  extremes.  With  further  investigation,  we  shall  probably  be  able  to 
devise  a  number  of  other  dichotomies,  so  that  our  classification  of  indi- 
viduals will  no  longer  be  a  matter  of  setting  them  in  a  row  between  two 
extremes.  With  every  new  dichotomy  a  new  dimension  would  have  to 
be  added  to  ooir  map  of  possible  variations. 

It  is  possible,  too,  that  some  of  our  future  subdivisions  will  occur  only 
among  schizothymes  or  only  among  cyclothymes,  and  indeed  it  already 
seems  to  me  that  a  number  of  different  syndromes  are  confused  under 
the  general  term  "schizothyme",  while  the  concept  of  cyclothymia  with 
its  good  correspondence  with  Jung's  "extravert"  and  Jaensch's  "inte- 
grate" is  more  likely  to  be  a  unity. 

Such  classification  is  an  important  preliminary  to  an  understanding 
of  the  working  of  the  various  systems  of  personality,  and  is  especially 
useful  in  presenting  problems  of  patterns  of  behaviour  between  per- 
sonalities. 


The  Preferred  Types  i6i 

ethos  and  Kretschmer's  cyclothymia,  I  must  insist  that  the 
resemblance  between  these  two  syndromes  is  solely  in  terms 
of  easy  emotional  acceptance  and  jolliness ;  and  that  I  know 
of  no  hint  of  periodic  variation  between  exuberant  joy  and 
depression  in  the  women's  ethos,  such  as  is  characteristic 
of  cycloid  personalities.  It  is  doubtful  whether  such  an 
individual  tendency  could  be  culturally  standardised  in  such 
a  way  that  all  the  individuals  were  "in  step"  with  one 
another,  i.e.  all  manic  at  one  time  and  all  depressed  at 
another.  More  probably,  if  periodic  changes  were  stan- 
dardised in  the  women,  each  would  have  her  separate  perio- 
dicity, and  it  would  be  hard  to  demonstrate  whether  this 
periodicity  was  a  product  of  cultural  standardisation  or  an 
expression  of  individual  physiological  deviance. 

My  material,  which  was  not  collected  with  these  problems 
in  view,  is  not  good  enough  to  be  the  basis  of  an  exact 
analysis  of  these  various  syndromes,  but  since  the  con- 
trasting sex  ethoses  of  latmul  culture  are  at  least  reminiscent 
of  the  types  described  by  Kretschmer,  it  is  worth  while  to 
consider  how  far  the  latmul  are  typologically  conscious.  To 
what  extent  have  they  developed  ideas  about  the  association 
of  physique  with  character,  and  what  types  do  they  recognise } 

The  natives  regard  two  types  of  man  with  approval.  The 
first  is  the  man  of  violence  and  the  second  the  man  of  dis- 
cretion. Of  these  the  violent  type  is  the  most  admired,  and 
such  a  man  is  described  with  enthusiasm  as  "  having  no  ears  ". 
He  pays  no  attention  to  what  is  said  to  restrain  him  but 
suddenly  and  recklessly  follows  his  assertive  impulses.  Such 
a  man  is  represented  in  Plate  XXII,  and  indeed  this  man  was 
a  little  too  sudden  and  unstable  even  for  latmul  taste.  They 
regarded  him  as  somewhat  "cranky"  and  warned  me  against 
him  when  I  took  him  as  an  informant.  In  this  capacity  he 
proved  more  curious  than  useful — very  enthusiastic,  but  too 
hasty  and  astonishingly  inaccurate.  He  seemed  indeed  to  lack 
all  power  of  critical  thought  and  to  have  no  sense  of  logical 
consistency.  When  his  contradictory  statements  were  pre- 
sented to  him  he  had  no  realisation  of  their  incompatibility.  ^ 

^  The  occurrence  of  such  a  cognitive  trait  in  a  man  who  was  the 
extreme  of  the  preferred  affective  type  is  especially  interesting  since  the 


1 62  The  Preferred  Types 

Such  men  though  admired  would,  I  was  told,  not  be 
trusted  with  esoteric  information,  because  the  natives  fear 
that  in  the  erudite  debating  about  the  system  of  names  and 
totems,  such  an  uncontrolled  person  may  blurt  out  some 
important  piece  of  secret  lore  or  provoke  a  brawl  by  too  rashly 
exposing  his  opponents'  secrets.  Thus  with  his  little  know- 
ledge of  esoterica,  the  violent  man  will  behave  in  debate  in 
the  sort  of  way  which  I  have  described  above,  filling  out  his 
speeches  with  histrionics  and  obscene  reference. 

The  more  discreet  type  is,  I  think,  generally  heavier — more 
pyknic — in  physique  and  quieter  and  rather  more  at  ease  in 
his  public  appearances.  He  it  is  who  is  the  repository  of 
mythological  knowledge,  and  it  is  he  who  contributes  erudi- 
tion to  the  totemic  debating  and  keeps  the  discussion  on 
more  or  less  systematic  lines.  His  balance  and  caution^  enable 
him  to  judge  whether  to  expose  his  opponents'  secrets  or 
merely  to  indicate  by  some  trifling  hint  that  he  knows  the 
secrets,  such  a  hint  being  tantamount  to  a  threat  of  exposure. 
He  knows  how  to  sit  quietly  in  the  debate  carefully  watching 
his  opponents  to  judge  whether  they  really  know  any  of  the 
important  secrets  of  his  clan  or  whether  their  trifling  hints 
are  only  a  bluff^  to  frighten  him  into  ceding  some  point. 

In  mythology  these  two  types  are  contrasted.  There  is  a 
series  of  tales  of  two  brothers  of  whom  the  elder,  Kamwaim- 
buangga,  was  of  the  discreet  type  while  the  younger,  Woli- 
ndambwi,  was  a  man  of  violence.  Of  these  it  is  the  latter 
who  is  the  great  hero,  but  who  in  fits  of  temper  set  fire  to 
the  original  mythological  ceremonial  house  and  killed  his 
sister's  son.  This  reckless  hero  is  said  to  have  been  a  man 
of  great  beauty  and  especially  to  have  had  a  long  nose  which 
was  much  admired.  It  was  also  stated  to  me  in  Mindimbit 
that  his  patrilineal  descendants,  the  members  of  the  half-clan 
which  claims  him  as  ancestor,  have  inherited  from  him  noses 
more  beautiful  than  those  of  the  descendants  of  his  discreet 

latmul  culture  itself  contains  so  many  "contradictory"  formulations. 
Questions  of  the  relationship  between  affective  and  cognitive  standardisa- 
tions will  be  considered  in  a  later  chapter. 

'  A  cyclothyme  observer  would  probably  prefer  to  describe  such  a  man 
as  "well  balanced",  while  to  a  schizothyme  he  appears  "cautious"  if  he 
be  a  friend,  or  "sly"  if  he  be  an  enemy. 


The  Preferred  Types  163 

elder  brother.  Among  these  beautiful  descendants  my  in- 
formant counted  Mwaim-nanggur,  the  mythological  hero 
whom  all  the  women  loved  and  who  was  finally  murdered 
by  their  jealous  husbands.  (Cf.  Oceania,  1932,  Plate  VI, 
which  shows  a  portrait  skull  of  Mwaim  with  enormous  nose.) 

I  was  for  a  long  time  puzzled  by  the  constant  references 
to  long  noses  and  by  the  conventional  exaggeration  of  the  nose 
in  artistic  representations  of  the  human  face  (Plate  XXVIII 
A  and  b).  But  the  matter  was  partly  cleared  up  by  a  discus- 
sion of  the  beautiful  woman's  head  represented  in  Plate  XXV. 
This  is  the  head  of  a  woman  who  died  young,  about  three 
generations  ago.  The  face  is  modelled  in  clay  upon  her  skull 
and  the  portrait  is  preserved  as  a  ritual  object  for  use  as  the 
head  of  a  mbwatnggowi  figure  in  certain  ceremonies  (cf. 
Plate  XXVII).  Among  the  heads  used  for  these  dolls,  some 
were  trophies  of  war,  while  others  were  the  heads  of  villagers 
who  had  died  in  peace.  I  asked  on  what  principle  her  head 
had  been  chosen,  and  the  natives  said  that  any  head  of  special 
beauty  might  be  picked  for  this  purpose  and  pointed  to  the 
length  of  the  leptorrhine  nose  as  the  conspicuously  attractive 
feature. 

Thus  the  exaggeration  of  the  nose  in  latmul  art  is  in  part 
at  least  a  conventional  reference  to  the  standard  of  beauty. 
But  there  is  probably  another  factor  at  work.^  The  enormous 
noses  are  certainly  to  some  extent  phallic  symbols.  They  may 
be  lengthened  downwards  to  join  the  penis  or  navel,  or  they 
may  end  free  with  a  representation  of  the  head  of  a  snake  or 
bird  at  the  tip.  I  have  indicated  elsewhere  that  in  this  culture 
phallic  symbols  are  to  be  regarded,  not  simply  as  symbols 
of  the  genital  organ,  nor  as  symbols  of  fertility,  but  rather 
as  symbols  of  the  whole  proud  ethos  of  the  males.  Thus  the 
flutes,  the  secrets  of  initiation,  are  phallic  in  mythological 
origin  and  stand  in  the  culture  as  an  outstanding  symbol  of 
the  differentiation  of  the  male  sex  by  drastic  initiation.  ^ 

^  I  collected  one  instance  only  in  which  a  small  nose  was  regarded 
as  desirable.  This  was  in  the  bullying  of  the  novices  by  a  masked  figure. 
Each  novice  was  made  to  rub  noses  with  the  mask  and  to  say:  "Your 
nose  is  a  small  nose,  mine  is  big  (nemwan) "  and  to  express  a  liking  for 
the  contact.   Otherwise,  I  invariably  heard  praise  for  long  {tshivla)  noses. 

*  "  Music  in  New  Guinea  ",  The  Eagle,  he.  cit. 


164  The  Preferred  Types 

In  ideas  about  noses,  the  symbolism  seems  to  have  gone 
full  circle.  The  schizothyme  ethos  of  the  men  is  linked  with 
emphasis  upon  a  leptorrhine  standard  of  beauty,  and  the  use 
of  phallic  symbols  for  the  schizothyme  ethos  has  involved 
the  use  of  the  leptorrhine  nose  as  a  phallic  symbol. 

Unfortunately  I  was  not  familiar  with  Kretschmer's  typo- 
logy when  I  was  in  the  field  and  therefore  did  not  enquire 
into  any  ideas  which  may  be  present  in  this  culture  as  to  the 
correlation  of  violent  character  with  leptorrhine  nose.  The 
attribution  of  such  a  nose  to  Woli-ndambwi  and  its  denial  to 
the  discreet  elder  brother,  however,  seem  to  indicate  that 
the  latmul  have  some  vague  idea  of  this  sort. 

It  is  interesting  that  the  same  type  of  nose  is  admired  in 
women,  in  whom  a  violent  proud  temperament  would  seem 
at  variance  with  their  jolly  co-operative  ethos.  But  in  spite 
of  the  contrast  in  sex  ethos  the  men  have  a  leptosome  standard 
of  beauty  for  both  sexes.  They  prefer  the  physical  type  which 
Kretschmer  maintains  is  associated  with  the  schizothyme 
temperament,  and  articulately  regret  that  latmul  women 
tend  to  have  "bad  small  noses". 

I  have  already  mentioned  that  the  men  feel  a  distaste  for 
the  women's  ethos,  and  it  seems  that  the  same  attitude  is 
reflected  in  their  standards  of  beauty.  But  while  it  is  certain 
that  the  men  admire  the  woman  who  recklessly  takes  the 
initiative  in  sexual  affairs,  even  risking  her  life  in  the  venture, 
it  is  by  no  means  certain  that  the  same  type  of  woman  is 
popular  with  her  own  sex.  I  did  not  collect  any  statements 
of  the  women's  views  on  these  matters.  It  is  possible  that 
the  leptorrhine  standard  of  beauty  in  women  is  linked  with 
the  occasional  emphasis  upon  pride  which  is  described  in  the 
chapter  on  the  women's  ethos. 

In  a  very  few  men  the  admired  qualities  of  both  the  violent 
and  discreet  types  were  combined,  or  so  it  seemed  to  me. 
Such  a  man  was  Mali-kindjin,  who  is  now  dead,  but  who  was 
the  greatest  man  in  Kankanamun.  Toravi  of  Angerman  has 
something  of  the  same  greatness. 

Mali-kindjin  was  an  old  man,  and  when  I  knew  him  he  was 
an  invalid.  They  said  that  his  sorcery  was  beginning  to  recoil 
upon  his  own  head.    He  was  an  astonishingly  vivid  and 


The  Preferred  Types  165 

dramatic  orator ;  when  one  entered  his  house  he  would  come 
forward  and  make  a  speech  of  welcome,  not  unctuous  but 
beautifully  crisp  and  hard.  In  the  initiatory  system,  he  was 
a  member  of  grade  2  (cf.  Diagram,  p.  245)  and  therefore 
nominally  had  no  active  part  in  the  business  of  initiation, 
but  he  was  always  on  the  spot  criticising  the  proceedings  and 
correcting  the  gong  rhythms.  On  one  occasion  the  father  of 
a  novice  was  angry  because  something  had  been  jabbed  into 
his  son's  eye.  Mali-kindjin  took  the  side  of  the  initiators  and 
in  the  debate  which  followed  he  suddenly  seized  a  log  from 
the  fire  on  the  ground  and  with  it  belaboured  the  angry 
father — who  took  his  punishment  in  silence. 

Shortly  before  he  died,  Mali-kindjin  was  trying  to  change 
his  own  position  in  the  initiatory  system,  to  get  himself 
demoted  so  that  he  might  take  part  in  the  ceremonial.  His 
father  had  caused  him  to  be  promoted  when  he  was  a  boy 
from  grade  6  to  grade  4  (cf.  Diagram,  p.  245).  He  now 
claimed  that  this  promotion  had  been  irregular,  and  that 
he  should  now  be  a  member,  not  of  grade  2,  but  of  grade  4, 
where  he  would  play  a  more  active  part. 

He  was  well  hated  and  feared  for  his  sorcery,  and  when  it 
was  clear  that  he  was  really  ill,  a  debate  was  organised  against 
him.  His  esoteric  claim  to  the  Sepik  River  as  a  clan  ancestor 
was  attacked.  He  told  the  members  of  his  own  clan  to  keep 
quiet  during  the  debate — he  would  do  all  the  speaking  him- 
self. After  four  hours  of  acrimonious  debate  and  insult,  con- 
tinually dancing  his  scorn  of  those  who  would  impugn  his 
Kindjin-kamboi  (the  snake  which  is  the  Sepik  River,  according 
to  his  clan's  mythology),  he  deliberately  exposed  an  esoteric 
secret  of  the  opposition.  He  picked  up  a  yam  tuber  and  began 
to  dance  with  it  in  his  mouth — a  reference  to  the  opposition's 
secret  myth  of  the  origin  of  flutes  from  a  yam.  A  brawl 
started  at  once  and  Mali-kindjin  got  some  nasty  knocks  and 
had  a  stool  thrown  at  his  head.  The  debate  was  resumed  and 
went  on  till  evening,  Mali-kindjin  still  keeping  up  his  flow 
of  oratory.  Then,  at  dusk,  when  it  was  all  over,  he  went  and 
sat  alone,  an  exhausted  and  sick  man,  at  the  end  of  the  cere- 
monial house,  and  began  to  chant  the  name  songs  of  his 
clan — his  eyes  shining. 


1 66  The  Preferred  Types 

The  feelings  of  the  other  natives  about  Mali-kindjin  were 
very  definitely  ambivalent.  For  example,  Tshava,  his  own 
sister's  child,  had  in  his  boyhood  a  great  deal  of  contact  with 
Mali-kindjin,  and  even  took  part  in  some  of  his  sorcery. 
Mali-kindjin  wanted  to  teach  Tshava  his  spells,  but  the  latter 
was  frightened.  He  says  that  he  was  frightened  lest  Mali- 
kindjin  should  take  offence  at  the  small  payment  for  the 
secrets,  but  I  suspect,  too,  that  he  was  frightened  of  the 
dangerous  position  in  which  a  sorcerer  finds  himself,  hated 
by  the  other  members  of  his  village.  In  any  case,  he  says 
that  it  was  partly  on  the  advice  of  his  father  that  he  did  not 
associate  more  with  Mali-kindjin,  and  now  he  regrets  this  and 
says:  "Now  me  stop  along  bloody  fool,  that's  all."  The 
"bloody  fool"  is  his  father. 

When  Tshava  discussed  Mali-kindjin's  doings  with  me,  he 
quite  clearly  admired  the  latter  for  his  breaches  of  convention. 
It  i6  usual  to  present  shell  valuables  to  the  medicine  man 
when  his  services  are  engaged:  but  these  valuables,  though 
nominally  they  become  the  property  of  the  medicine  man, 
are  usually  returned  to  the  donor.  The  patient  when  he  is 
cured  brings  to  the  medicine  man  a  bunch  of  areca  nuts  and 
places  these  upon  the  latter's  shoulder.  The  medicine  man 
then  returns  the  valuables  which  had  been  given  to  him. 
Tshava  told  me  with  glee  that  Mali-kindjin  always  kept  both 
the  valuables  and  the  areca  nuts. 

Mali-kindjin's  son,  Avuran-mali,  had  also  a  very  great  ad- 
miration for  the  old  sorcerer  and  pride  in  his  doings.  After 
the  debate  mentioned  above  in  which  Mali-kindjin  was  badly 
knocked  about,  Avuran  boasted  to  me  of  how  his  father 
had  won  the  debate — how  he  had  exposed  his  opponents' 
secrets  and  thereby  exasperated  them  into  a  brawl. 

Other  people,  however,  not  so  closely  related  to  the  old 
sorcerer,  regarded  him  with  frank  detestation.  It  was  pointed 
out  to  me  as  evidence  of  his  sorcery  that  "he  had  no  rela- 
tives ",  the  inference  being  that  he  had  killed  his  own  relatives 
to  prevent  the  sorcery  from  recoiling  upon  his  own  head. .  But 
though  they  feared  him,  even  hated  him,  they  were  proud, 
too,  that  he  was  a  member  of  their  village,  and  recognised 
that  Mali-kindjin  in  the  old  days  had  helped  them  both 
magically  and  physically  in  warfare. 


The  Preferred  Types  167 

Another  type  of  man  who  is  recognisable  in  every  village 
is  the  sorcerer  who  lacks  violence — skinny,  pinched  indi- 
viduals who  have  compensated  for  their  lack  of  spirit  by 
cultivating  a  reputation  for  skill  in  magic.  Such  a  man  was 
Namwio  of  Mindimbit,  a  henpecked  husband  for  whose  wife 
everyone  was  sorry ;  and  Tshaun-awan  of  Palimbai,  a  man  who 
had  never  married,  was  another  of  this  type.  Such  men  are 
useful ;  they  perform  cures  and,  as  public  servants,  carry  out 
the  magic  to  cause  the  water  level  to  rise  and  fall  when 
required.  They  also  may  be  paid  to  cause  death  and  sickness 
or  to  raise  the  water  level  in  the  interests  of  some  individual 
at  a  time  when  the  majority  would  benefit  by  a  fall.  I  think 
that  in  the  past  little  or  nothing  was  done  to  punish  these 
activities,  but  nowadays  there  is  a  tendency  to  try  to  persuade 
the  District  Officer  to  put  these  men  in  prison  as  scapegoats 
for  the  vagaries  of  the  floods.  As  informants  I  have  come  in 
contact  with  three  of  these  men,  and  in  each  case  I  have  been 
surprised  to  find  that  their  esoteric  knowledge  is  actually  very 
limited.  They  cannot  really  recite  their  spells,  but  only 
jumble  them.  The  native  comment  upon  these  men  is — 
"they  have  no  meat  on  their  bones." 

Another  insult  which  is  levelled  at  personalities  whom  the 
natives  do  not  admire  is  the  phrase:  ''kau  tapman  nyan" 
(literally  "fight — none — child",  i.e.  a  man  with  no  fight  in 
him). 

There  is  another  line  of  enquiry  into  native  personalities 
which  I  did  not  pursue  but  which,  I  believe,  would  have 
given  positive  results.  Though  my  material  is  inconclusive, 
it  is  perhaps  worth  giving  since  it  presents  a  new  aspect  of 
the  problems  of  culture  contact.  At  the  present  time,  the 
villages  of  the  latmul  contain  considerable  numbers  of  young 
men  who  have  recently  returned  to  their  homes  after  spending 
from  three  to  five  years  as  indentured  labourers  on  European 
plantations  and  gold  mines.  They  grew  up  as  boys  in  the 
latmul  ethos  probably  admiring  it  and  probably  believing 
that  the  ethos  was  "natural"  for  men.  Then  they  went  away 
and  lived  for  some  years  in  the  more  disciplined  and  co- 
operative ethos  of  a  labour  line.  Now  they  are  returning  to 
the  ethos  in  which  they  grew  up. 


1 68  The  Preferred  Types 

I  believe  that  among  these  young  men,  when  I  recall  their 
personalities,  I  can  roughly  discriminate  those  to  whom  the 
schizothyme  behaviour  was  natural  or  innate,  and  that  such 
young  men  when  they  come  home  drop  easily  back  into  the 
native  community.  But  the  others  appear  to  have  been  bom 
with  or  to  have  acquired  but  little  bent  for  schizothyme 
behaviour.  As  boys,  they  probably  adopted  the  prevailing 
ethos  because  it  was  fashionable  and  was  the  only  pattern 
offered  to  them.  But  now,  when  they  have  seen  and  lived 
in  a  different  ethos,  these  men  look  askance  at  their  native 
culture.  They  are  impatient  of  the  buffooning  of  the  older 
men  and  they  treat  the  most  important  rituals  with  contempt. 
They  are  openly  careless  about  the  secrets  of  initiation. 

An  anecdote  will  illustrate  the  sort  of  heresy  which  is  now 
growing  in  the  community  as  a  result  of  the  culture  contact : 

The  men  ceremonially  concluded  an  elaborate  wagan  cere- 
mony which  had  lasted  for  many  weeks.  During  all  that  time^ 
the  wagan,  the  sacred  initiatory  slit-gongs,  had  been  beaten 
and  there  had  been  a  strict  taboo  on  the  use  of  the  secular 
gongs.  At  the  close  of  the  ceremony,  the  men,  in  ritual  pro- 
cession, put  away  the  sacred  heads  of  the  two  wagan  which 
had  been  exhibited  in  front  of  the  gongs  during  the  ceremony 
and  then  they  beat  a  special  rhythm  on  the  secular  gongs, 
thus  removing  the  taboo. 

In  the  very  early  morning  of  the  following  day,  I  was 
woken  by  some  men  who  came  to  my  bedside  and  spoke 
somewhat  as  follows:  "We  want  to  kill  Tshimbat's  pig. 
Formerly  we  would  have  killed  him,  but  now  we  are  afraid 
of  prison.  If  we  kill  his  pig  will  the  Government  put  us  in 
prison?  Yesterday  he  and  the  other  boys  in  the  toy  cere- 
monial house  carried  out  a  mock  ceremony  in  imitation  of 
ours.  They  beat  the  rhythm  before  we  did.  They  made  our 
ceremony  futile." 

Tshimbat  (Plate  XXIV  b)  was  a  very  bumptious,  noisy, 
somewhat  pyknic,  returned  labourer,  whose  house  the  men 
had  had  to  raid  some  months  previously  at  dead  of  night, 
prodding  the  house  with  long  bamboos  to  overturn  and 
smash  the  pots,  because  Tshimbat  had  been  fooling  with  a 
^  Cf.  footnote,  p.  137. 


The  Preferred  Types  169 

hand-drum  when  he  ought  not.  I  knew  him  well  as  a  disturber 
of  the  peace  and  knew  that  the  Government  would  support 
what  was  fine  in  the  native  culture.  I  said:  "My  shoot  boy 
shall  take  my  gun  and  shoot  the  pig."  This  was  done  and 
we  all  ate  it. 

That  day  Tshimbat's  mother  came  out  onto  the  dancing 
ground  and  stood  just  in  front  of  the  ceremonial  house, 
dirging  for  the  pig  which  she  had  raised.  A  fine  schizothyme 
old  man,  Djuai  (Tshava's  father),  seized  a  stick  and  beat  her 
thoroughly,  driving  her  off. 

But  the  dice  are  heavily  loaded  in  favour  of  the  discontented 
and  maladjusted  young  men,  and  a  further  interference  from 
me  was  necessary  before  my  conservative  friends  were  finally 
successful.  A  few  days  later,  entering  the  ceremonial  house 
one  morning,  I  was  surprised  to  see  a  row  of  currency  shells 
set  out  as  if  a  subscription  was  being  raised  for  some  feast. 
I  asked  what  feast  was  in  the  air,  and  was  told : "  No.  Tshimbat 
has  been  talking.  He  is  going  to  make  a  court  case  of  it. 
We  shall  all  go  to  prison.  He  knows  pidgin  English  and  we 
don't.  We  are  going  to  pay  for  the  pig."  Then  I  made  a 
speech  at  the  debating  stool  and  said  that  the  only  man  who 
was  legally  liable  was  Djuai  and  that  the  shells  should  be 
put  in  a  basket  and  given  to  him  if  he  suffered ;  that  the 
pig  had  been  killed  by  my  shoot  boy  with  my  gun  and  acting 
on  my  orders  and  that  under  no  circumstances  would  I  give 
anything  to  Tshimbat.  So  the  shells  were  put  in  a  string 
bag,  and  we  all  waited  for  the  District  Officer. 

When  the  case  was  heard,  Tshimbat  spoke  querulously  of 
the  old  men,  stated  a  preference  for  the  customs  of  the  white 
men,  and  received  a  very  sharp  snub  for  his  pains.  Djuai 
said  little,  but  stood  puzzled,  nervous  and  rather  stiff.  He 
was  fined  ten  shillings  for  assault.  No  damages  were  awarded 
to  Tshimbat  and  his  mother. 

Then  we  withdrew  to  the  ceremonial  house  to  consider 
what  should  be  done  with  the  bag  of  currency  shells.  The 
natives  dislike  changing  shells  into  shillings  and  it  was  there- 
fore decided  that  the  shells  should  be  returned  to  the  donors 
and,  without  interference  from  me,  a  fresh  collection  of 
shillings  was  started.    I  contributed  one  shilling,  and  four 


lyo  The  Preferred  Types 

more  were  added.  Then  Djuai  came  forward  to  the  debating 
stool.  He  said  that  he  had  wanted  for  other  reasons  to  hit 
that  woman.  She  had  been  unkind  to  his  child,  and  he  would 
be  well  content  with  five  shillings. 

In  this  instance  I  was  able  to  defend  the  culture,  but 
without  my  interference  it  is  almost  certain  that  the  initiatory 
system  in  Kankanamun  would  have  received  a  severe  blow. 
I  suggest  that  a  very  great  deal  of  the  action  of  culture  contact 
in  destroying  institutions  may  be  due  to  mechanisms  of  this 
kind — the  contact  upsetting  the  delicate  adjustment  between 
the  temperaments  of  deviant  individuals  and  the  ethos  of  the 
culture. 


CHAPTER  XIII 

Ethological  Contrast^  Competition 
and  Schismogenesis 

THE  foregoing  description  of  the  contrast  between  the 
men's  and  women's  ethos  among  the  latmul  people 
raises  at  once  the  problem  of  how  such  ethological 
contrast  is  produced  and  maintained.  The  first  theory  which 
occurs  to  us  is  that  individuals  of  the  two  sexes,  owing  to 
genetic  and  deep  physiological  differences,  may  tend  to  de- 
velop different  patterns  of  personality,  and  that  the  differences 
in  ethos  in  the  two  sexes  are  simply  an  expression  of  these 
innate  differences.  But  though  we  invoke  heredity  we  cannot 
exclude  the  influence  of  culture  and  environment. 

The  matter  is  confused  and  difficult.  The  two  extreme 
theories,  (a)  that  the  ethological  contrast  is  entirely  deter- 
mined by  culture  and  (6)  that  the  contrast  is  entirely  deter- 
mined by  sex  physiology  and  heredity,  are  both  of  them 
untenable.  The  first  must  be  abandoned  because  we  know 
of  considerable  innate  physical  and  physiological  differences 
between  the  sexes,  differences  in  shape,  in  bulk,  in  deposition 
of  fat,  in  speed  of  development  and  so  on.  These  differences 
must  necessarily  be  expressed  in  all  the  behaviour  of  the 
individuals — in  gesture,  posture,  choice  of  activity,  etc. — and 
therefore  must  contribute  to  the  ethological  contrast. 

On  the  other  hand,  the  theory  which  would  ascribe  the 
contrast  entirely  to  heredity  must  be  abandoned  for  various 
reasons.  Even  if  we  were  dealing  with  physical  differences 
between  the  sexes  in  any  one  community,  the  question  would 
arise  as  to  how  much  these  physical  differences  are  culturally 
exaggerated  or  modified  by  costume  or  deformation ;  and  the 
likelihood  of  cultural  modification  of  psychological  differences 
is  even  greater. 

Some  theory  of  a  fundamental  biological  difference  in 
temperament  between  the  sexes  may  probably  be  found  in 
every  community  in  which  the  ethoses  of  the  two  sexes  are 


172  Ethological  Contrast,  Competition 

differentiated,  and  in  our  own  culture  with  its  extreme  con- 
fusion and  diversity  of  ethos,  it  is  usual  to  ascribe  almost  any 
characteristic  of  a  personality  to  biological  sex.  Under  the 
term  "womanliness"  is  subsumed  a  syndrome  of  charac- 
teristics— warm  motherliness,  easy  emotional  expression, 
pyknic  beauty  and  so  on — comparable  with  the  ethos  of 
women  among  the  latmul.  But  the  term  "  femininity  "  is  used 
for  a  syndrome  much  more  closely  allied  to  latmul-masculine- 
pride,  sudden  capriciousness,  waywardness  and  leptosome 
beauty. 

More  significant,  however,  than  the  confused  phrasings  of 
our  own  culture  are  the  findings  of  Dr  Margaret  Mead.^  She 
has  shown  that  among  the  Arapesh  who  live  in  the  mountains 
north  of  the  Sepik  River,  between  it  and  the  coast,  a  uniform 
ethos  is  standardised  for  both  sexes;  and  that  among  the 
Mundugumor  who  live  on  a  tributary  of  the  Lower  Sepik, 
there  is  again  no  contrast  in  sex  ethos.  But  while  the  ethos 
of  both  sexes  among  the  Arapesh,  though  more  gentle,  is 
vaguely  reminiscent  of  that  which  we  have  found  among  the 
latmul  women,  the  ethos  of  both  sexes  among  the  Mundugu- 
mor is  a  harsher  and  less  exhibitionistic  version  of  that  of  the 
latmul  men.  Lastly,  Dr  Mead  worked  among  the  Tchambuli 
and  found  there  a  contrast  between  the  ethos  of  the  two  sexes. 
This  contrast  was,  however,  not  identical  with  that  which 
I  have  described  among  the  latmul.  The  ethos  of  the  Tcham- 
buli men  was  less  harsh  and  more  exhibitionistic  than  that 
of  the  latmul,  while  the  Tchambuli  women  were  somewhat 
harder  and  more  business-like  than  those  of  the  latmul.  In 
the  light  of  these  findings,  if  we  are  to  maintain  a  theory  of 
innate  difference  in  temperament  in  the  two  sexes,  we  should 
have  to  suppose  that  the  genetic  constitution  of  the  indi- 
viduals is,  statistically,  markedly  different  in  each  of  the  four 
tribes  which  I  have  mentioned. 

Inasmuch  as  there  are  physical  differences  between  these 
peoples  and  a  likelihood  that  these  physical  differences  are 
genetically  determined,  it  is  conceivable  that  there  may  be 
differences  in  innate  temperament.  But  at  least  we  can  say 
that  the  differences  are  not  due  to  peculiarities  of  sexual 

^  Sex  and  Temperament,  1935. 


and  Schismogenesis  173 

physiology,  since  either  ethos  may  be  standardised  for  both 
sexes.  The  differences,  whether  sexual  or  not,  would  have 
to  be  described  in  Mendelian  terms,  and  I  know  of  no  pattern 
of  Mendelian  inheritance  which  would  enable  us  to  build 
up  populations  with  prominent  statistical  differences  or  simi- 
larities between  the  sexes  such  as  occur  in  the  four  tribes 
which  have  been  described. 

We  are  thus  forced  to  take  up  some  position  intermediate 
between  the  two  extreme  theories.  .We  cannot  entirely  ex- 
clude either  heredity  or  social  environment,  and  it  would  be 
premature  to  indicate  the  exact  form  of  intermediate  theory 
which  we  should  adopt.  In  this  chapter,  since  the  material 
I  collected  was  not  of  a  sort  to  be  susceptible  of  genetic 
analysis,  and  since  I  took  no  measurements  of* physique, 
I  must  confine  myself  to  the  investigation  of  the  social  and 
cultural  factors  involved  in  the  shaping  of  personality,  while 
leaving  open  the  possibility  that  such  temperamental  charac- 
teristics as  are  referred  to  in  the  terms,  cyclothymia  and 
schizothymia,  may  be  determined  by  heredity  and  are  prob- 
ably independent  of  sex. 

It  is  possible,  for  example,  that  the  populations  in  the  four 
tribes  mentioned  above  are  statistically  alike  in  tempera- 
mental characteristics,  and  that  in  each  population  and  in 
each  sex  there  occur  individuals  who  are  born  with  a  natural 
bent  for  the  various  ethoses.  If  this  could  be  demonstrated, 
we  should  then  have  to  suppose  that  some  of  the  individuals 
in  any  one  culture  are  naturally  more  fitted  for  life  in  that 
culture  than  others,  and  that  in  each  culture  there  are  genetic 
deviants  who  do  their  best  to  adapt  themselves  to  an  ethos 
whose  bias  is  but  little  developed  in  their  temperaments.  We 
should,  however,  still  have  to  acknowledge  that  hereditary 
tendencies  had  played  a  very  important  part  in  the  shaping 
of  the  culture  since  the  innate  characteristics  of  the  "pre- 
ferred types  "  would  seem  to  have  guided  the  culture  in  its 
evolution. 

With  this  preambk,  we  may  turn  to  consider  cultural 
factors  which  probably  promote  the  ethos  of  each  sex  in 
latmul  culture.  When  I  was  among  the  latrhul,  the  very 
concept  of  ethos  was  to  me  very  dimly  defined,  and  I  have 


1 74  Ethological  Contrast^  Competition 

never  had  any  training  or  practical  experience  in  the  study 
of  the  moulding  of  human  beings  and  especially  children. 
I  have  therefore  not  the  material  to  describe  these  processes 
among  the  latmul  in  any  detail,  and  my  remarks  must  be 
regarded  as  no  more  than  a  tentative  suggestion  of  how  the 
processes  may  fit  together. 

My  impression  is  that  there  is  no  marked  difference  be- 
tween the  treatment  of  male  babies  and  that  of  females,  nor 
did  I  find  any  strong  feeling  that  babies  of  one  sex  were  more 
desirable  than  those  of  the  other.  In  general,  infants  of  both 
sexes  seem  to  be  happy  and  well-treated ;  they  are,  I  think, 
rarely  left  alone  for  long  periods.  One  detail  is  worth  men- 
tioning :  in  this  culture  children  of  both  sexes  are  conspicu- 
ously ornamented  with  shells  even  in  everyday  life  (cf. 
Plate  XVII  a).i  I  used  to  carry  on  my  wrist  a  circular  pig's 
tusk  (bought  in  Sydney)  and  found  that  the  loan  of  this  shiny 
white  object,  obviously  an  ornament,  was  invariably  eff"ective 
in  stopping  the  crying  of  small  children  of  both  sexes,  at  least 
for  a  few  minutes. 

It  is  my  impression  that  we  should  look  for  the  origins  of 
contrasting  sex  ethos  among  the  latmul,  not  in  the  experiences 
of  very  early  childhood,  but  in  the  later  training  of  boys  and 
girls.  We  should  see  the  two  ethoses  as  acquired  by  learning 
and  imitation  rather  than  as  springing  from  peculiarities 
implanted  in  the  deep  unconscious  in  the  first  two  years  of 
life. 

In  the  case  of  the  boys :  the  preoccupation  of  their  seniors 
with  head-hunting  and  with  the  production  of  spectacular 
displays;  their  life  in  the  toy  ceremonial  house,  where  they 
ape  their  seniors;  the  processes  which  we  have  analysed  in 
our  study  of  initiation  (p.  130  et  seq.);  the  little  boy's  first 
experience  of  homicide  when  he  is  still  a  child,  spearing  some 
wretched  bound  captive,  while  his  wau  helps  him  to  lift 
the  spear;  the  elaborate  naven,  of  which  the  little  boy  is 
immediately  made  the  hero — all  these  factors,  no  doubt, 
contribute  to  shape  the  boys  into  latmul  men. 

^  It  is  possible  that  this  ornamentation  of  children  is  in  some  sense 
analogous  with  the  ornamentation  of  women,  which  is  a  conspicuous 
characteristic  of  the  Eastern  latmul  (cf.  footnote,  p.  142). 


and  Schismogenesis  175 

Similarly,  the  women's  ethos  is  no  doubt  formed  in  part 
by  their  preoccupation  with  the  routines  of  food-getting  and 
child-rearing,  and  by  the  association  of  girls  with  older 
women  who  have  already  adopted  the  ethos. 

Factors  of  the  sort  which  I  have  considered  above  may  be 
presumed  to  play  their  part  in  maintaining  the  status  quo, 
but  in  addition  to  these  I  believe  that  in  the  mechanisms 
which  underlie  ethological  contrast,  we  are  concerned  with 
other  factors  which,  if  they  were  unrestrained,  would  lead 
to  changes  in  the  cultural  norms.  I  am  inclined  to  see  the 
Status  quo  as  a  dynamic  equilibrium,  in  which  changes  are 
I  continually  taking  place.  On  the  one  hand,  processes  of 
differentiation  tending  towards  increase  of  the  ethological 
contrast,  and  on  the  other,  processes  which  continually 
counteract  this  tendency  towards  differentiation. 

The  processes  of  differentiation  I  have  referred  to  as 
schismogenesis.^  They  are,  I  believe,  of  very  wide  sociological 
and  psychological  significance,  and  therefore  in  my  descrip- 
tion of  these  phenomena  I  shall  use  as  illustrative  material 
not  only  the  very  meagre  facts  which  I  collected  in  New 
Guinea,  upon  which  the  concept  of  schismogenesis  was 
originally  built,  but  also  no  less  sketchy  observations  of  the 
occurrence  of  schismogenesis  in  European  communities. 
r  I  would  define  schismogenesis  as  a  process  of  differentiation 
in  the  norms  of  individual  behaviour  resulting  from  cumulative 
interaction  between  individuals. 

For  the  moment  we  need  not  concern  ourselves  with 
defining  the  exact  position  of  this  concept  in  regard  to  the 
various  disciplines  which  I  have  tried  to  separate.  I  think 
that  we  should  be  prepared  rather  to  study  schismogenesis 
from  all  the  points  of  view — structural,  ethological,  and  socio- 
logical— which  I  have  advocated ;  and  in  addition  to  these  it 
is  reasonably  certain  that  schismogenesis  plays  an  important 
part  in  the  moulding  of  individuals.  I  am  inclined  to  regard 
the  study  of  the  reactions  of  individuals  to  the  reactions  of  other 
individuals  as  a  useful  definition  of  the  whole  discipline  which 

^  Man,  1935,  p.  199,  "Culture  Contact  and  Schismogenesis".  This 
article  is  an  outline  of  the  sociological  problems  presented  by  the  concept 
of  schismogenesis.  It  contains  almost  no  reference  to  the  phenomena 
which  I  observed  in  New  Guinea. 


176  Ethological  Contrast^  Competition 

is  vaguely  referred  to  as  Social  Psychology.  This  definition 
might  steer  the  subject  away  from  mysticism. 

We  should  do  well,  I  think,  to  speak  no  more  of  "the  social 
behaviour  of  individuals"  or  of  "the  reactions  of  the  indi- 
vidual to  society".  These  phrasings  lead  all  too  easily  to 
such  concepts  as  those  of  Group  Mind  and  Collective  Uncon- 
scious. These  concepts  are  almost  meaningless  to  me,  and 
I  believe  that  even  if  we  avoid  them  we  are  liable  to  err  by 
confusing  our  study  of  the  psychological  processes  of  the 
individual  with  our  study  of  society  as  a  whole ;  a  confusion 
of  spheres  of  relevance. 

When  our  discipline  is  defined  in  terms  of  the  reactions 
of  an  individual  to  the  reactions  of  other  individuals,  it  is 
at  once  apparent  that  we  must  regard  the  relationship  between 
two  individuals  as  liable  to  alter  from  time  to  time,  even 
without  disturbance  from  outside.  We  have  to  consider,  not 
only  A's  reactions  to  5's  behaviour,  but  we  must  go  on  to 
consider  how  these  affect  5's  later  behaviour  and  the  effect 
of  this  on  A. 

It  is  at  once  apparent  that  many  systems  of  relationship, 
either  between  individuals  or  groups  of  individuals,  contain 
a  tendency  towards  progressive  change.  If,  for  example,  one 
of  the  patterns  of  cultural  behaviour,  considered  appropriate 
in  individual  A,  is  culturally  labelled  as  an  assertive  pattern, 
while  B  is  expected  to  reply  to  this  with  what  is  culturally 
regarded  as  submission,  it  is  likely  that  this  submission  will 
encourage  a  further  assertion,  and  that  this  assertion  will 
demand  still  further  submission.  We  have  thus  a  potentially 
progressive  state  of  affairs,  and  unless  other  factors  are  pre- 
sent to  restrain  the  excesses  of  assertive  and  submissive 
behaviour,  A  must  necessarily  become  more  and  more  asser- 
tive, while  B  will  become  more  and  more  submissive;  and 
this  progressive  change  will  occur  whether  A  and  B  are 
separate  individuals  or  members  of  complementary  groups. 

Progressive  changes  of  this  sort  we  may  describe  as  com- 
plementary schismogenesis.  But  there  is  another  pattern  of 
relationships  between  individuals  or  groups  of  individuals 
which  equally  contains  the  germs  of  progressive  change.  If, 
for  example,  we  find  boasting  as  the  cultural  pattern  of 


and  Schismogenesis  177 

behaviour  in  one  group,  and  that  the  other  group  repHes  to 
this  with  boasting,  a  competitive  situation  may  develop  in 
which  boasting  leads  to  more  boasting,  and  so  on.  This  type 
of  progressive  change  we  may  call  symmetrical  schismo- 
genesis.^ 

In  the  light  of  this  theoretical  consideration  of  the  possi- 
bilities, it  is  clear  that  we  must  consider  the  various  contexts 
of  latmul  culture  to  determine  whether  either  complementary 
or  symmetrical  schismogenesis  contributes  anything  to  the 
shaping  of  cultural  norms.  Is  it  possible  that  the  contrast 
in  sex  ethos  is  of  such  a  kind  that  it  is  liable  to  complementary 
schismogenesis  ? 

We  have  seen  that  the  women  are  an  audience  for  the 
spectacular  performances  of  the  men,  and  there  can  be  no 
reasonable  doubt  that  the  presence  of  an  audience  is  a  very 
important  factor  in  shaping  the  men's  behaviour.  In  fact, 
it  is  probable  that  the  men  are  more  exhibitionistic  because 
the  women  admire  their  performances.  Conversely,  there  can 
be  no  doubt  that  the  spectacular  behaviour  is  a  stimulus 
which  summons  the  audience  together,  promoting  in  the 
women  the  appropriate  complementary  behaviour.  We  may 
wonder,  too,  whether  the  whole  system  of  behaviour  which 
surrounds  the  flutes,  wagan,  and  other  secrets  of  initiation 
would  be  maintained  if  it  were  not  for  the  fact  that  the  women 
hear  and  admire  the  music  of  the  flutes  and  the  rhythms  of 
the  wagan. 

But  the  contrast  between  exhibitionism  and  admiration  is 
only  a  part  of  the  general  sex  contrast  which  includes  a 
whole  nexus  of  inter-related  characteristics,  and  it  would  be 
very  important  to  know  whether  this  wider  contrast  tends 
towards  schismogenesis.  The  only  detail  which  I  can  pro- 
duce which  seems  to  show  that  such  a  schismogenesis  occurs 
in  latmul  culture  is  the  reaction  of  the  men  to  the  dirging 

^  The  difference  between  complementary  and  symmetrical  schismo- 
genesis is  closely  analogous  to  that  between  schism  and  heresy,  where 
heresy  is  the  term  used  for  the  splitting  of  a  religious  sect  in  which  the 
divergent  group  have  doctrines  antagonistic  to  those  of  the  parent  group, 
while  schism  is  the  term  used  for  the  splitting  of  a  sect  in  which  two 
resulting  groups  have  the  same  doctrine,  but  separate  and  competing 
politics.  In  spite  of  this,  I  have  used  the  term  schismogenesis  for  both 
types  of  phenomena. 


178  Ethological  Contrast^  Competition 

of  the  widow,  when  they  indulge  in  harsh  caricature.  Unfor- 
tunately I  do  not  know  the  widow's  reaction  to  being  carica- 
tured. 

Complementary  schismogenesis  is  evident  again  in  the 
contexts  of  initiation,  the  process  of  inculcating  latmul- 
masculine  ethos  into  the  novices.  This  process  I  have  already 
analysed  in  such  detail  as  the  material  permits  (p.  133). 
I  have  indicated  how  the  initiators  with  latmul-masculine 
ethos  react  to  the  presence  of  novices  with  (supposedly) 
latmul-womanly  ethos:  the  presence  of  the  novices  stimu- 
lates the  initiators  to  harsh  and  fantastic  behaviour.  Later, 
the  novices  themselves  become  initiators,  and  are  schismo- 
genically  driven  by  the  presence  of  later  novices  into  latmul- 
masculine  behaviour.  The  admiring  ethos  of  the  women  is 
also  active  in  bringing  about  the  assimilation  of  the  novices 
to  the  group  of  men;  the  novices  are  ornamented  and  ex- 
hibited to  the  women  at  the  end  of  initiation. 

Lastly,  a  curious  and  possibly  factitious  case  of  comple- 
mentary schismogenesis  may  be  noted  in  the  laua's  boasting 
in  the  presence  of  his  wau,  who  shames  himself  in  reply. 

Symmetrical  schismogenesis  is  not  evident  between  the 
sexes,  but  occurs  in  dramatic  form  in  initiation.  Here  we 
have  rival  moieties  competing  against  each  other  in  their 
bullying  of  the  novices,  and  prompting  each  other  to  further 
brutalities  (pp.  134,  135). 

The  phenomena  of  Schismogenesis  are  by  no  means  con- 
fined to  latmul  culture,  and  in  order  to  emphasise  the  wide- 
spread importance  of  the  process  which  I  first  noted  among 
the  latmul,  I  shall  indicate  in  what  other  fields  I  expect  to 
recognise  schismogenesis : 

I.  In  all  intimate  relations  between  pairs  of  individuals. 
A  great  many  of  the  maladjustments  of  marriage  are  nowadays 
described  in  terms  of  the  identification  of  spouse  with  parent. 
Such  a  phrasing  may  be  historically  accurate :  it  may  be  true 
that  the  husband  in  a  marriage  tends  to  carry  over  into  his 
relationship  with  his  wife  attitudes  which  have  been  pre- 
viously formed  in  his  relationship  with  his  mother.  But  this 
fact  alone  is  by  no  means  sufficient  to  account  for  the  break- 


and  Schismogenesis  C  179] 

down  of  the  marriage,  and  it  is  difficult  in  terms  of  such  a 
theory  to  explain  why  such  marriages,  in  their  earlier  stages, 
are  often  very  satisfactory  and  only  later  become  a  cause  of 
misery  to  both  partners. 

But  if  we  add  to  this  diachronic  phrasing  of  the  relation- 
ship the  possibility  that  the  patterns  of  behaviour  between 
the  partners  are  liable  to  progressive  change  of  a  schismo- 
genic  nature,  it  is  evident  that  we  have  a  theory  which  would 
explain  both  why  the  relationship  is  satisfactory  in  its  early 
stages  and  why  its  breakdown  appears  inevitable  to  the  people 
concerned.  The  relationship  between  son  and  mother  is,  in  our 
culture,  a  complementary  relationship,^  which  in  its  early 
stages  is  patterned  on  fostering  on  the  mother's  side  and 
feebleness  on  the  child's.  Later  the  relationship  may  develop 
in  very  various  ways:  e.g.,  {a)  the  pattern  of  fostering  and 
feebleness  may  persist;  or  {b)  the  relationship  may  evolve 
towards  the  pattern  which  we  have  noted  among  the  latmul 
where  the  mother  takes  a  vicarious  pride  in  her  offspring; 
or  {c)  it  may  evolve  towards  an  assertive-submissive  contrast 
in  which  it  may  be  either  person  who  takes  the  assertive  role. 
But  whatever  the  pattern,  the  mother-son  relationship  is 
almost  always  complementary.  If  these  patterns  are  carried 
over  into  the  son's  marriage  it  is  likely  that  they  may  there 
become  the  starting  point  of  a  schismogenesis  which  will 
wreck  the  marriage. 

It  is  perhaps  worth  while  to  suggest  that  in  many  cases  an 
explanation  and  demonstration,  to  the  partners  in  such  mar- 
riages, of  the  schismogenesis  in  which  they  are  involved 
might  have  the  same  therapeutic  effect  as  the  understanding, 
on  the  part  of  the  husband,  that  he  is  identifying  his  wife 
with  his  mother. 

2.  In  the  progressive  maladjustment  of  neurotic  and 
prepsychotic  individuals.  I  myself  have  no  experience  of 
psychiatry,  but  I  suspect  that  in  addition  to  studying  the 

^  Exceptionally,  in  our  own  culture,  we  may  come  across  cases  in  which 
the  relationship  between  son  and  mother  is  almost  symmetrical  or  reci- 
procal. It  would  be  interesting  to  know  whether  such  sons  are  ever, 
in  their  marriages,  involved  in.  complementary  schismogenesis  of  the 
festering-feebleness  or  exhibitionism-admiration  types. 


i8o,.  Ethological  Contrast^  Competition 

individual  pathology  in  every  case,  the  psychiatrist  would  do 
well  to  pay  more  attention  to  the  relations  which  the  deviant 
individual  has  with  those  around  him.  I  have  discussed  this 
matter  with  Dr  J.  T.  MacCurdy,  who  agrees  with  me  that 
in  many  cases  the  growth  of  the  symptoms  of  the  paranoid 
individual  are  attributable  to  schismogenic  relationships  with 
those  nearest  to  him.  I  understand  that  it  is  usual  to  find 
that  those  paranoids  who  build  their  delusions  around  a 
belief  in  the  unfaithfulness  of  their  wives,  almost  invariably 
have  wives  whose  utter  faithfulness  is  obvious  to  every  out- 
sider. Here  we  may  suspect  that  the  schismogenesis  takes 
the  form  of  continual  expression  of  anxiety  and  suspicion  on 
the  husband's  side,  and  continual  response  to  this  on  the  side 
of  the  wife,  so  that  she,  either  continually  humouring  him  or 
contradicting  him,  is  promoting  his  maladjustment,  and  he, 
in  turn,  becoming  more  maladjust,  demands  more  and  more 
exaggerated  responses  from  her. 

In  the  case  of  schizoid  maladjustment,  the  matter  is  not 
so  clear.  I  have  suggested  above  that  in  latmul  culture  the 
circular  ethos  of  the  women  and  the  schizothyme  ethos  of 
the  men  are  mutually  complementary  and  liable  to  schismo- 
genesis. If  this  be  true,  and  further  observations  are  required 
to  verify  it,  we  must  be  prepared  to  accept  the  fact  that  the 
schizophrene  is  not  merely  working  out  his  own  internal 
pathology,  which  indeed  may  or  may  not  be  getting  worse, 
but  is  also  responding  to  the  more  cyclothyme  people  around 
him  by  himself  becoming  more  and  more  schizoid. 

Such  a  view  of  the  progressive  degeneration  of  the  schizoid 
would  explain  very  simply  the  extraordinary  inevitability  of 
this  degeneration,  and  would  explain,  too,  the  fact  that  the 
patient  himself  is  often  preoccupied  with  ideas  of  predestina- 
tion in  such  a  degree  that  these  ideas  themselves  contribute 
in  no  small  measure  to  his  destruction.  It  is  possible  that  if 
in  the  early  stages  a  schizoid  individual  could  be  brought  to 
realise  that  the  progressive  process  which  he  sees  as  inevitable 
is  actually  very  simple  and  easily  controlled,  its  progress 
might  be  arrested. 

But  a  great  deal  of  work  remains  to  be  done  in  labelling 
the  various  syndromes  which  are  grouped  under  the  heading 


and  Schismogenesis  i8i 

of  "schizothymia".  From  the  latmul  material,  it  is  clear 
only  that  the  contrast  between  the  men  and  the  women  is 
somehow  comparable  to  that  between  one  type  of  schizo- 
thymia  and  the  circular  temperament.  It  is  probable,  how- 
ever, that  there  are  a  good  many  patterns  of  complementary 
schismogenesis  in  each  of  which  one  partner  is  ultimately 
driven  into  schizoid  distortions  of  his  personality,  and  that  a 
classification  of  these  various  paired  patterns  will  give  us  a 
clue  to  the  classification  of  the  schizothyme  syndromes.  More- 
over, an  understanding  of  these  syndromes  in  terms  of  the 
schismogeneses  in  which  they  arise  may  give  us  a  clue  to 
their  treatment  in  the  early  stages. 

In  the  later  stages  of  schizophrenia  it  seems  probable  that 
the  personality  of  the  patient  is  permanently  maimed,  and 
an  understanding  of  the  schismogenesis  which  contributed 
to  his  breakdown  would  probably  be  of  no  therapeutic  use. 

It  is  probable  that  schismogenesis  is  an  important  factor 
in  neurosis  as  well  as  in  psychosis,  and  that  a  new  discipline 
of  psycho-analysis  could  be  built  up  on  these  lines  supple- 
menting the  systems  which  are  now  being  used.  In  Freudian 
analysis  and  in  the  other  systems  which  have  grown  out  of  it, 
there  is  an  emphasis  upon  the  diachronic  view  of  the  indi- 
vidual, and  to  a  very  great  extent  cure  depends  upon  inducing 
the  patient  to  see  his  life  in  these  terms.  He  is  made  to  realise 
that  his  present  misery  is  an  outcome  of  events  which  took 
place  long  ago,  and,  accepting  this,  he  may  discard  his  misery 
as  irrelevantly  caused.  But  it  should  also  be  possible  to  make 
the  patient  see  his  reactions  to  those  around  him  in  synchronic 
terms,  so  that  he  would  realise  and  be  able  to  control  the 
schismogenesis  between  himself  and  his  friends. 

It  is  likely  that  while  for  some  patients  the  administering 
of  a  diachronic  view  is  curative  in  its  effects,  for  others  this 
treatment  may  only  accentuate  their  maladjustment.  For  these 
latter,  it  is  possible  that  the  administering  of  a  synchronic 
view  would  be  curative  and  give  them  a  complete  and  realistic 
understanding  of  themselves.  The  dangers  of  psycho-analysis 
when  administered  to  schizoid  patients  may  indeed  arise 
simply  out  of  the  preoccupation  of  such  patients  with  destiny 
and  the  inevitability  of  historic  accident.    A  sense  of  con- 


1 82  Ethological  Contrast,  Competition 

temporary  process  is  perhaps  a  necessary  corrective  to  an 
over-developed  sense  of  personal  history,  and  mce  versa. 

I  believe  also  that  it  would  be  worth  while  to  look  for 
schismogenic  phenomena  in  cases  of  split  personality.  Here 
we  have  two  forms  of  schismogenesis  to  consider:  first  a 
probable  schismogenesis  between  the  patient  and  his  friends, 
and  secondly  a  possible  schismogenesis  within  the  pefsonality 
of  the  patient.  It  is  possible,  in  fact,  that  one  half  of  the  split 
personality  promotes  the  other  and  vice  versa,  producing  an 
ever-widening  breech  and  incidentally  causing  each  half  of 
the  personality  to  be  less  and  less  capable  of  adaptive  behaviour 
in  the  patient's  social  setting. 

To  what  extent  this  schismogenesis  within  the  personality 
is  present  in  all  individuals  who  are  involved  in  an  external 
schismogenesis  is  a  question  which  cannot  now  be  answered, 
but  it  is  probable  that  we  may  see  the  development  of 
vicarious  pride  and  its  expression  in  gestures  of  self-abnega- 
tion as  phenomena  related  to  those  of  split  personality.  It  is 
probable  too  that  we  should  see  exhibitionism  and  narcissism 
as  very  closely  inter-related  phenomena.  The  latmul  man, 
who  courts  the  admiration  of  the  women  by  exhibitionistic 
behaviour,  is  almost  certainly  admiring  his  own  performance 
with  one  half  of  his  personality :  an  external  schismogenesis 
may  induce  an  internal  narcissism. 

Dr  MacCurdy  has  suggested  two  other  contexts  in  which 
schismogenesis  probably  occurs,  namely  in  the  behaviour  of 
manic  patients,  and  in  the  tantrums  of  children.  The  symptoms 
of  mania  are  enormously  increased  when  an  audience  is  pre- 
sent; and  the  responses  of  parents  are  liable  to  accentuate 
the  tantrums  of  their  children. 

It  is  not  enough,  however,  to  make  a  list  of  contexts  in 
which  schismogenesis  occurs,  and  it  would  be  a  mistake  to 
suppose  that  in  the  word  schismogenesis  we  have  the  key  to 
all  the  processes  of  character  formation.  If  this  concept  is 
to  be  of  real  value,  it  must  stimulate  enquiry  into  the  con- 
ditions upon  which  schismogenesis  depends.  Among  these 
we  may  note  that  schismogenesis  is  impossible  unless  the 
social  circumstances  are  such  that  the  individuals  concerned 
are  held  together  by  some  form  of  common  interest,  mutual 


and  Schismogenesis  183 

dependence,  or  by  their  social  status.  As  to  the  nature  of 
the  factors  which  hold  them  together,  and  the  relationship 
between  these  factors  and  the  schismogenic  process,  we  know 
nothing. 

Another  factor  which  is  necessary  for  schismogenesis  has 
already  been  mentioned  but  it  assumes  a  special  importance 
in  these  psychological  contexts.  I  said  (p.  176)  that  if  the 
behaviour  oi  A  "is  culturally  labelled  as  an  assertive  pattern, 
while  B  is  expected  to  reply  to  this  with  what  is  culturally 
regarded  as  submission",  we  may  expect  schismogenesis  to 
occur.  The  ethological  aspect  of  the  behaviour  is  fundamental 
for  schismogenesis,  and  we  have  to  consider  not  so  much  the 
content  of  the  behaviour  as  the  emotional  emphasis  with  which 
it  is  endowed  in  its  cultural  setting. 

If  A  gives  B  some  object,  and  B  receives  it,  these  acts  may 
be  seen,  according  to  their  setting,  as:  {a)  a  triumph  for  A, 
an  attitude  which  may  lead  either  to  symmetrical  schismo- 
genesis of  the  "potlatch"  type  or  to  some  form  of  comple- 
mentary schismogenesis  if  the  asymmetry  is  continued ;  [b)  a 
triumph  for  B,  an  attitude  which  may  lead  to  some  schismo- 
genesis on  the  general  lines  of  commercial  rivalry,  or  (c)  a 
triumph  for  neither.  Both  giving  and  receiving  may  be  seen 
merely  as  expressions  of  mutual  friendliness,  and  their  con- 
tinuance may  lead,  not  to  schismogenesis,  but  rather  to  closer 
union  between  A  and  B. 

When  we  are  dealing  with  schismogenesis  between  two 
separate  individuals  or  groups  of  individuals,  it  is  clear  that 
we  must  consider  ethological  emphasis  as  a  necessary  con- 
dition for  schismogenesis.  But  the  question  becomes  more 
difficult  when  we  consider  schismogenesis  within  a  single 
personality.  It  is  probable  that  some  condition  analogous  to 
"ethological  emphasis"  must  be  postulated  for  such  internal 
schismogeneses,  but  it  is  not  clear  how  this  is  to  be  phrased 
without  unduly  personifying  the  separate  elements  in  the 
multiple  personality:  a  problem  which  must  be  left  to  the 
psychiatrists. 

3,  In  culture  contacts.  I  have  already  given  an  outline  of 
how  considerations  derived  from  the  concept  of  schismo- 


184  Ethological  Contrast^  Competition 

genesis  should  be  applied  to  the  study  of  culture  contact.^ 
In  my  opinion,  we  should  see  the  phenomena  of  contact  as 
a  series  of  steps  starting  from  a  point  at  which  two  groups 
of  individuals,  with  entirely  different  cultures  in  each  group, 
come  into  contact.  The  process  may  end  in  various  ways,  and 
the  theoretically  possible  end-results  of  the  process  may  be 
enumerated:  (i)  the  complete  fusion  of  the  two  groups, 
(2)  the  elimination  of  one  or  both  groups,  (3)  the  persistence 
of  both  groups  in  dynamic  equilibrium  as  differentiated 
groups  in  a  single  major  community. 

At  the  beginning  of  the  contact,  however,  at  least  where  the 
contact  is  one  between  Europeans  and  "primitive"  people, 
we  cannot  expect  to  find  any  simple  relationship  between  the 
ethos  of  the  one  group  and  that  of  the  other.  But  I  suspect 
that  in  quite  a  short  time  the  individuals  of  each  group  adopt 
special  norms  of  behaviour  in  their  contacts  with  individuals 
of  the  other  group,  and  that  these  special  norms  of  behaviour 
will  be  classifiable  in  terms  of  complementary  or  symmetrical 
patterns. 

This  phenomenon  is  clearly  seen  in  New  Guinea  in  the 
behaviour  which  Europeans  adopt  towards  natives,  and  in 
that  which  the  natives  adopt  towards  Europeans.  Several 
hundreds  of  different  cultures,  with  the  greatest  diversity  of 
ethos  among  them,  have  been  concerned  in  this  culture  con- 
tact, and  yet  one  fact  strikes  the  anthropologist  as  soon  as  he 
reaches  the  country:  the  average  European  resident  believes 
that  the  natives  of  New  Guinea  are  remarkably  similar  in  all 
parts  of  the  country.  The  resident  will  grant  that  some  are 
better  workers,  and  that  others  have  more  courage,  but 
broadly  his  opinion  is  that  one  "coon"  is  very  much  like 
another.  The  basis  for  this  opinion  is,  I  believe,  to  be  found 
in  the  special  behaviour  which  the  natives  adopt  in  their 
dealings  with  Europeans.  Apart  from  slight  differences,  the 
majority  of  peoples  of  Mandated  New  Guinea  have  adopted 
the  same  tactics  in  dealing  with  Europeans,  and  they  have 
no  doubt  copied  each  other  in  their  methods. 

Pidgin  English,  the  language  which  has  developed  in  this 
culture  contact,  and  the  system  of  gestures  and  intonation 

^  Man,  193s,  p.  199. 


and  Schismogenesis  185 

which  are  almost  a  part  of  pidgin,  seem  to  me  to  be  an 
expression  of  a  very  definite  ethos.  The  language  has  its  own 
cadence  and  special  flexibility,  which  we  may  suppose  adapted 
to  humour  the  sometimes  assertive  and  always  incalculable 
European.  I  am  inclined  even  to  think  that  the  ethos  ex- 
pressed in  pidgin  English  is  comparable  with  that  which  is 
caricatured  in  American  "  negro-humour  ".^ 

In  any  case,  it  is  certain  that  for  a  majority  of  the  natives 
of  New  Guinea  the  ethos  of  pidgin  contrasts  sharply  with 
the  ethos  of  their  native  cultures.  This  was  conspicuously  true 
of  the  latmul,  and  the  behaviour  of  returned  labourers  in  the 
ceremonial  house,  etc.,  was  a  continual  source  of  shock  to 
me.  I  felt  that  it  was  vulgar  while  the  behaviour  of  the  other 
natives,  though  it  might  be  harsh  and  noisy,  had  never  this 
peculiar  quality.  I  think  that  the  natives  who  had  never 
worked  on  plantations  felt  shocked  in  the  same  way. 

The  process,  however,  by  which  two  groups  whose  respec- 
tive cultures  are  mutually  irrelevant  evolve  a  complementary 
or  symmetrical  relationship  in  terms  of  behaviour  which  is 
normal  to  neither  group,  has  never  been  investigated.  It 
should,  if  susceptible  of  study,  give  us  clues  to  the  process 
of  establishment  of  schismogenic  pairs. 

Once  complementary  patterns  of  behaviour  have  been 
established,  it  is  my  opinion  that  subsequent  schismogenesis 
is  responsible  for  many  of  the  antipathies  and  misunder- 
standings which  occur  between  groups  in  contact.  Hand  in 
hand  with  the  ethological  divergence,  we  find  the  develop- 
ment of  structural  premises  which  give  permanence  and 
fixity  to  the  split.  But  it  is  not  clear  to  what  extent  these 
formulations,  of  "colour-bar",  "racial  antipathy",  and  mu- 
tual avoidance,  contribute  to  promoting  the  schismogenesis. 
It  is  possible  that  some  of  these  formulations  prevent  it 

^  Dr  MacCurdy  has  pointed  out  to  me  an  interesting  detail  in  the 
schismogenesis  between  negro  and  white  in  America — that  the  Southern 
negroes  in  a  highly  differentiated  schismogenic  relationship  with  the 
Southern  whites  are  very  unwilling  to  accept  from  Northern  whites 
behaviour  patterns  they  would  expect  and  demand  from  the  whites 
whom  they  know.  Thus  the  patterns  of  behaviour  developed  in  a  schismo- 
genesis cannot  easily  be  applied  in  other  contexts,  and  for  an  under- 
standing of  the  equilibrium  we  shall  need  to  know  a  great  deal  more  than 
the  mere  details  of  schismogenic  behaviour. 


1 86  Ethological  Contrast^  Competition 

from  going  too  far.  These  are  problems  which  need  in- 
vestigation. 

Ultimately  the  two  groups  reach  one  of  the  end  results 
which  I  have  listed  above.  Of  these,  we  may  ignore  the  first 
two  and  consider  only  the  persistence  of  both  groups  in 
dynamic  equilibrium,  a  relationship  between  the  groups  in 
which  the  tendencies  towards  schismogenesis  are  adequately 
restrained  or  counteracted  by  other  social  processes.  This  end 
result  is  closely  analogous  to  the  state  of  affairs  which  exists 
in  a  primitive  community  between  any  two  differentiated 
groups,  and  it  is  likely  that  a  study  of  such  equilibria  as  those 
between  the  initiatory  moieties  or  between  the  sexes  in  latmul 
society  will  contribute  to  our  knowledge  of  these  equilibria, 
and  will  enable  us  to  handle  the  problems  of  culture  contact 
in  a  more  understanding  manner. 

4.  In  politics.  In  the  present  disturbed  and  unstable  state 
of  politics  in  Europe,  there  are  two  schismogeneses  which 
stand  out  conspicuously :  (a)  the  symmetrical  schismogenesis 
in  international  rivalries,  and  {h)  the  complementary  schismo- 
genesis in  "class-war".  Here  again,  as  in  the  other  contexts 
in  which  we  have  discussed  schismogenesis,  the  progressive 
evolution  of  behaviour  towards  greater  and  greater  differen- 
tiation and  mutual  opposition  is  evident,  and  our  politicians 
would  appear  to  be  as  incapable  of  handling  the  process  as 
the  schizoid  is  of  adjusting  himself  to  reality. 

A  third  type  of  schismogenesis  is  of  some  local  importance, 
namely  the  process  whereby  dictators  are  pushed  towards  a 
state  which,  in  the  eyes  of  the  world,  seems  almost  psycho- 
pathic. This  is  a  complementary  schismogenesis  between  the 
dictator  on  the  one  hand,  and  his  officials  and  people  on  the 
other.  It  illustrates  very  clearly  how  the  megalomaniac  or  para- 
noid forces  others  to  respond  to  his  condition,  and  so  is  auto- 
matically pushed  to  more  and  more  extreme  maladjustment. 

These  vast  and  confused  schismogeneses  are  played  out  amid 
such  complicated  circumstances  that  it  is  scarcely  possible 
to  study  them.  But  it  might  conceivably  be  worth  while 
to  observe  to  what  extent  in  their  policies  politicians  are 
reacting  to  the  reactions  of  their  opponents,  and  to  what 


and  Schismogenesis  187 

extent  they  are  paying  attention  to  the  conditions  which  they 
are  supposedly  trying  to  adjust.  It  may  be  that  when  the 
processes  of  schismogenesis  have  been  studied  in  other  and 
simpler  fields,  the  conclusions  from  this  study  may  prove 
applicable  in  politics. 

THE  PROGRESS  AND  CONTROL 
OF  SCHISMOGENESIS 

In  the  absence  of  any  carefully  collected  material  which 
would  show  the  progress  of  schismogenesis  in  any  one  case, 
I  can  only  offer  my  opinion  as  to  what  stages  are  likely  to 
occur.  In  the  first  instance,  when  the  complementary  or 
symmetrical  relationship  is  first  established,  it  is  probable 
that  the  patterns  of  behaviour  which  the  two  individuals  or 
two  groups  adopt  seem  to  both  parties  to  be  a  satisfactory 
answer  to  a  difficult  problem  of  relationship.  Such  at  least 
would  seem  to  be  the  case,  for  example,  in  the  adoption  of 
pidgin  English  ethos  by  the  natives  of  New  Guinea,  and 
probably  the  same  is  true  of  the  early  stages  of  schismogenic 
marriage.  In  the  case  of  psychological  maladjustments,  how- 
ever, I  must  leave  it  to  the  psychiatrists  to  decide  whether 
the  initial  adoption  of  the  complementary  pattern  is  an  adap- 
tive process, 
f  If,  however,  a  complementary  or  symmetrical  pattern, 
when  once  adopted,  becomes  more  and  more  emphasised  by 
schismogenesis,  it  is  likely  that  the  personalities  of  the  indi- 
viduals concerned  will  undergo  some  sort  of  distortion  with 
over-specialisation  in  some  one  direction,  whether  it  be  ex- 
hibitionism, fostering,  assertion,  submission  or  what  not. 
With  this  distortion  some  degree  of  discomfort  will  be 
introduced  into  the  relationship,  and  it  may  even  happen, 
though  this  requires  verification,  that  the  individuals  in  trying 
to  find  again  the  answer  which  was  formerly  satisfactory, 
actually  specialise  even  further  in  their  respective  roles. 

Sooner  or  later,  the  distortion  of  the  personalities  is  likely 
to  be  accompanied  by  three  effects :  {a)  a  hostility  in  which 
each  party  resents  the  other  as  the  cause  of  its  own  distortion, 
(b)  at  least  in  complementary  schismogenesis,  an  increasing 


1 88  Ethological  Contrast,  Competition 

inability  to  understand  the  emotional  reactions  of  the  other 
party,  and  (c)  mutual  jealousy. 

At  a  comparatively  late  stage  in  the  schismogenesis,  when 
the  personalities  of  the  members  of  the  two  groups  have 
definitely  begun  to  suffer  from  distortion,  it  is  probably  usual 
to  find  a  development  of  mutual  envy.  The  distortion  is  2, 
progressive  specialisation  in  certain  directions  and  results  In 
a  corresponding  under-development  of  other  sides  of  the 
personality.  Thus  the  members  of  each  group  see  the  stunted 
parts  of  their  own  affective  life  fully  developed — indeed  over- 
developed— in  the  members  of  the  opposite  group.  It  is  in 
such  situations  that  mutual  envy  arises.  Not  only  do  the 
serfs  envy  the  aristocrats,  but  the  latter  develop  a  distaste 
for  their  own  ethos  and  begin  to  crave  the  simple  life. 

It  is  not  easy  to  judge  how  much  of  this  mutual  envy  is 
developed  between  the  sexes  in  latmul  culture.  Certainly  the 
women  take  a  very  real  pleasure  in  the  adoption  of  male 
clothes  and  ways  in  the  naven,  and  this  factor — mild  envy 
of  the  masculine  ethos — may  be  regarded  as  an  important 
motive  in  determining  their  transvesticism.  But  in  the  men, 
the  corresponding  envy  cannot  be  detected.  Outwardly,  at 
least,  they  despise  the  womanly  ethos,  but  it  is  not  impossible 
that  they  have  some  unacknowledged  envy  of  it.  Their  own 
ethos  would  not,  in  any  case,  permit  them  to  acknowledge 
that  there  was  anything  to  be  said  for  the  attitudes  of  the 
women  and  any  envy  which  they  may  have  they  might  well 
express  in  scorn  of  women — dressing  in  women's  clothes  for 
the  purpose.^ 

^  I  am  here  indebted  to  Dr  Karin  Stephen,  who  on  the  basis  of  her 
psycho-analytic  experience  in  Europe  suggested  this  possibility  to  me — 
that  unconscious  jealousy  of  the  women's  ethos  may  be  one  of  the  motives 
of  the  men's  transvesticism. 

I  am  very  doubtful  whether  such  a  phrasing  of  the  matter  is  permissible. 
I  have  endeavoured  to  use  references  to  emotions  only  where  such 
phrasing  could  be  regarded  as  a  rough,  makeshift  description  of  the 
behaviour  of  individuals  (cf.  footnote,  p.  124)  and  I  hesitate  to  launch 
into  phrasings  which  would  render  my  "behaviouristic"  references  to 
emotion  ambiguous  by  implying  that  the  behaviour  may  be  the  reverse 
of  that  primarily  appropriate  to  the  emotion. 

I  grant  in  general  that  theories  of  inversion,  etc.,  are  an  attempt  to 
express  an  important  truth,  but  I  suspect  that  the  theories  cannot  be 
properly  formulated  or  proven  till  we  have  techniques  for  analytical 
description  of  gestures,  posture,  etc.,  and  can  define  the  differences 
between  inverted  and  direct  behaviour. 


and  Schismogenesis  189 

It  is  likely  that  the  further  apart  the  personalities  evolve 
and  the  more  specialised  they  become,  the  more  difficult  it 
will  be  for  them  each  to  see  the  other's  point  of  view.  Finally, 
a  point  is  reached  where  the  reactions  of  each  party  are  no 
longer  a  striving  after  the  answer  which  was  formerly  satis- 
factory, but  are  simply  the  expression  of  distaste  for  the  type 
of  emotional  adjustment  into  which  the  other  party  has  been 
forced.  The  personalities  thus  become  mutually  contra-sug- 
gestible. In  place  of  patterns  of  behaviour  which  were  perhaps 
originally  adopted  in  an  attempt  to  fit  in  with  the  other  party, 
we  now  have  patterns  of  behaviour  which  are  definitely  a 
reaction  against  the  other  party.  Thus  the  schismogenesis 
takes  a  new  form,  and  the  relationship  becomes  less  and  less 
stable. 

If  we  see  schismogenesis  as  a  process  in  which  each  party 
reacts  to  the  reactions  of  the  other,  then  it  follows  that  the 
differentiation  must,  other  things  being  equal,  proceed  ac- 
cording to  some  simple  mathematical  law;  even  that  the 
specialisation  of  each  party  in  its  particular  patterns  of 
behaviour  should  follow  an  exponential  curve  if  plotted 
against  time.  We  cannot,  however,  hope  to  get  accurate 
measurements  of  the  progress  of  the  phenomenon  in  the 
majority  of  cases  of  schismogenesis,  but  it  is  possible  that 
in  some  relatively  isolated  case  some  detail  of  behaviour  will 
serve  as  an  index  of  degree  of  specialisation,  and  we  shall  be 
able  to  judge  what  sort  of  curve  occurs  in  that  case. 

In  general,  however,  not  only  is  the  specialisation  not 
susceptible  of  measurement,  but  also  the  progress  of  the 
schismogenesis  is  likely  to  be  modified  by  the  most  various 
factors.  We  may  mention  here  two  types  of  factors  which 
may  be  supposed  to  precipitate  and  hasten  the  process : 

{a)  An  individual  living  in  a  culture  is  trained  to  regard 
certain  patterns  of  behaviour  as  commendable,  and  certain 
others  as  wrong.  If,  in  a  schismogenesis,  he  is  led  to  overstep 
the  limits  of  what  is  culturally  approved,  such  of  his  behaviour 
as  passes  these  limits  is  likely  to  have  disproportionate  effects, 
not  only  in  making  him  conscious  of  the  distortion  of  his 
personality,  but  also  in  driving  the  other  party  to  over-drastic 
replies.    But  against  this,  it  must  be  remembered  that  the 


190  Ethological  Contrast^  Competition 

cultural  taboos  upon  certain  sorts  of  behaviour  may  be  active 
in  restraining  the  schismogenesis. 

(6)  A  second  factor  which  may  well  hasten  the  process  is 
a  tendency  to  believe  that  its  outcome  will  inevitably  be 
tragic.  The  schizoid  individual's  preoccupation  with  destiny 
may  well  be  not  only  a  realisation  that  natural  process  has 
got  him  in  its  clutches ;  it  may  be  a  factor  which  hastens  the 
process  towards  its  final  outcome. 

The  picture  of  schismogenesis  which  I  have  so  far  given 
is  of  a  process  inevitably  advancing  towards  such  differentia- 
tion that  some  outside  factor  is  bound  to  precipitate  the  final 
collapse.  I  have  described  the  schizoid  as  if  his  degeneration 
into  schizophrenia  was  inevitable.  I  have  described  the  break- 
down of  marriage  in  such  terms  that  it  is  inexplicable  that  any 
marriage  should  ever  be  a  success.  I  have  described  the 
difficulties  of  culture  contact  in  terms  which  would  imply 
that  the  final  equilibrium  which  I  mentioned  could  never  be 
achieved.  But  we  know  that,  in  actual  practice,  schizoid 
individuals  may  maintain  the  same  degree  of  maladjustment 
for  long  periods.  We  know  that  marriages  are  sometimes  suc- 
cessful, and  that  they  sometimes  recover  from  an  unstable 
condition.  We  must  therefore  think  of  schismogenesis,  not 
as  a  process  which  goes  inevitably  forward,  but  rather  as  a 
process  of  change  which  is  in  some  cases  either  controlled  or 
continually  counteracted  by  inverse  processes. 

The  term  "dynamic  equilibrium"  which  I  have  applied 
to  these  balanced  states  is  borrowed  from  the  jargon  of 
chemistry:  it  is  there  used  to  describe  apparently  stationary 
equilibria  which  theoretically  can  only  be  described  in  terms 
of  two  or  more  opposed  chemical  reactions  occurring  simul- 
taneously. It  is  in  a  sense  closely  comparable  with  this  that 
I  use  the  term  in  the  present  context.  The  analogy  between 
the  dynamic  equilibria  which  occur  in  human  society  and 
those  of  chemistry  cannot,  of  course,  be  pushed  very  far; 
but  it  should  serve  a  useful  purpose  in  reminding  us  that  the 
position  of  equilibrium  may  vary  enormously  from  one 
schismogenesis  to  another  and  from  one  culture  to  another. 
A  condition  of  mutual  opposition  which  would  spell  disaster 
for  many  English  marriages  may  well  be  an  equilibrium 


and  Schismogenesis  191 

position  for  marriages  in  other  cultural  settings;  and  the 
equilibrium  reached  in  the  sex  differentiation  among  the 
latmul,  with  its  steep  ethological  contrasts,  is  very  different 
from  the  corresponding  differentiation  in  England.  The 
rivalry  between  initiatory  moieties  in  latmul  society  would 
seem  to  be  stable,  but  the  state  of  affairs  which  obtains 
between  these  moieties  is  one  in  which  not  only  ritual  combat 
but  also  continual  brawls  are  culturally  normal. 

We  must  expect  also  to  find  a  very  different  strength  of 
restraining  factors  in  different  contexts  of  the  same  culture. 
I  have  mentioned  above  that  the  complementary  patterns 
which  are  culturally  normal  in  a  relationship  between  mother 
and  son  may  become,  in  our  culture,  the  basis  of  serious 
schismogenesis  when  they  are  carried  over  into  the  relationship 
between  man  and  wife.  We  must  ask  why  these  patterns  are 
dangerous  in  one  context  and  safe  in  another;  and  it  is 
probable  that  we  must  look  for  the  answer  either  in  the 
presence  of  adequate  restraining  factors  in  the  one  case  and 
their  absence  in  the  other,  or  in  the  fact  that  these  comple- 
mentary behaviour  patterns  are  culturally  expected  and  there- 
fore easily  tolerated  in  one  relationship,  while  they  are  in 
some  degree  culturally  disparaged  in  another. 

We  have  as  yet  no  carefully  collected  material  which  will 
demonstrate  the  existence  of  such  processes,  but  from  our 
general  knowledge  of  human  behaviour,  it  is  possible  to  make 
a  number  of  suggestions  as  to  the  nature  of  the  factors  which 
restrain  schismogenesis.  These  factors  may  be  classified  into 
two  main  groups.  We  began  this  chapter  with  a  consideration 
of  the  processes  which,  by  moulding  each  individual  to  his 
appropriate  ethos,  tend  to  maintain  the  status  quo.  From  this 
we  went  on  to  consider  schismogenesis  as  a  process  which, 
if  it  were  permitted  to  run  its  course,  would  bring  about 
changes  in  the  status  quo.  Here  again,  in  considering  the 
factors  which  control  schismogenesis,  we  have  to  deal  with 
some  whose  only  effect  is  to  limit  the  progress  of  the  schismo- 
genesis, and  with  others  which,  if  they  were  permitted  to  run 
their  course,  would  bring  about  changes  in  the  status  quo. 

In  the  first  category,  the  factors  that  we  have  to  consider 
are  processes  of  the  moulding  of  the  individuals.  While  the 


192  Ethological  Contrast,  Competition 

latmul  boy  is  trained  to  admire  harshness  and  himself  to  be 
harsh,  he  is  also  trained  to  regard  certain  extremes  of  harsh 
behaviour  as  reprehensible.  More  especially  he  learns  that 
such  extremes  are  dangerous,  and  that  others  would  take 
vengeance  upon  him  if  he  put  them  into  practice.  There  is 
thus  an  upper  limit  of  tolerance  of  assertive  behaviour,  and 
it  is  likely  that  a  similar  limit  of  tolerance  exists  in  the  case 
of  exhibitionism. 

In  my  account  of  latmul  ethos,  I  have  said  little  of  its 
upper  limits.  To  a  European,  the  emphases  of  latmul  mascu- 
line ethos  are  so  striking  that  he  does  not  easily  realise  that 
the  ethos  is  not  only  promoted  to  the  point  of  harshness 
which  he  observes,  but  is  also  prevented  from  developing 
beyond  that  point.  There  are,  however,  indications  in  the 
material  which  I  collected  which  show  the  reality  of  this 
upper  limit  of  tolerance.  I  have  mentioned  that  the  individual 
whose  photograph  (Plate  XXII)  gives  the  best  impression  of 
latmul  masculine  ethos  was  actually  rather  too  unstable  for 
native  taste.  Another  example,  which  illustrates  the  position 
of  the  man  who  carries  latmul  ethos  too  far,  may  be  seen  in 
the  story  of  Mwaim-nanggur  {Oceania,  ^932,  Plate  VI).  This 
hero  had  a  beautiful  nose,  and  he  originated  the  process  of 
extracting  sago ;  but  all  the  women  loved  him,  and,  with  the 
detached  humour  typical  of  the  latmul,  he  arranged  them 
in  two  rows,  those  with  hair  on  their  pubes  on  one  side  and 
those  without  on  the  other :  he  then  copulated  with  them  all. 
At  this  point,  the  limiting  mechanisms  came  into  play,  and 
Mwaim-nanggur  was  murdered  by  the  husbands  of  the  women. 

Even  success  in  head-hunting  may  be  carried  too  far,  and 
it  is  still  remembered  against  Malenembuk,  who  died  about 
four  generations  ago  in  Palimbai,  that  when  he  was  on  the 
prow  of  the  war  canoe  he  always  killed  and  left  no  one  for 
others  to  kill. 

In  general,  while  the  latmul  culture  emphasises  pride  and 
harshness,  channels  are  provided  for  the  legitimate  expression 
of  this  ethos,  and  if  a  man  specialises  too  far  in  violence,  his 
wives  will  run  away  from  his  house,  his  brothers-in-law  will 
turn  against  him,  and  he  will  live  under  a  threat  of  perhaps 
violent  death,  but  certainly  sorcery. 


and  Schismogenesis  193 

Among  the  processes  which,  like  schismogenesis,  would  be 
progressive  in  their  unrestrained  action,  but  which  may  be 
balanced  against  the  latter  process,  we  may  count  the  fol- 
lowing : 

1 .  It  is  possible  that  actually  no  healthy  equilibrated  rela- 
tionship between  two  groups  or  two  individuals  is  either 
purely  symmetrical  or  purely  complementary,  but  that  every 
such  relationship  contains  elements  of  the  other  type.  It  is 
certainly  easy  to  classify  relationships  into  one  or  the  other 
category  according  to  their  predominant  emphasis,  but  it  is 
possible  that  a  very  small  admixture  of  complementary 
behaviour  in  a  symmetrical  relationship,  or  a  very  small 
admixture  of  symmetrical  behaviour  in  a  complementary 
relationship,  may  go  a  long  way  towards  stabilising  the  posi- 
tion. For  example,  the  Squire  is  in  a  predominantly  comple- 
mentary and  not  always  comfortable  relationship  with  his 
villagers,  but  if  he  participates  in  village  cricket  (symmetrical 
rivalry)  but  once  a  year,  this  may  have  a  curiously  dispro- 
portionate effect  in  easing  the  schismogenic  strain  in  the 
relationship. 

In  latmul  culture,  we  have  noted  a  number  of  small  incon- 
sistencies in  the  ethos  of  each  sex.  In  the  case  of  the  men, 
we  noted  that  two  types  are  singled  out  for  approbation :  first 
and  foremost  the  man  of  violence,  but  after  him  the  man 
of  discretion.  It  is  possible  that  the  trust  and  liking  which 
other  individuals  have  for  the  discreet  man  may  XVell  con- 
tribute an  undercurrent  of  almost  latmul-womanly  ethos,  in 
terms  of  which  some  of  the  patterns  of  behaviour  between 
the  sexes  may  be  symmetrical.  Among  the  women,  we  have 
seen  even  more  clearly  a  double  emphasis  running  through 
their  ethos;  while  in  their  everyday  life  they  are  jolly  and 
co-operative,  upon  ceremonial  occasions  they  tend  towards 
the  proud  ethos  of  the  men.  It  is  certain  that  they  are  then 
admired  by  the  men,  and  these  occasions  may  well  be  of  con- 
siderable importance  in  reducing  the  opposition  between  the 
sexes. 

2.  It  is  possible  that  where  the  principal  emphasis  in  a 
schismogenic  relationship  is  upon  one  complementary  pair 
of  patterns,  that  the  admixture  of  complementary  patterns 


194  Ethological  Contrast^  Competition 

from  another  potentially  schismogenic  pair  may  stop  the  first 
schismogenesis.  To  give  a  hypothetical  case,  it  is  likely  that, 
in  a  marriage  in  which  there  is  a  complementary  schismo- 
genesis based  upon  assertion-submission,  illness  or  acci- 
dent may  shift  the  contrast  to  one  based  on  fostering  and 
feebleness.  Such  a  shift  may  bring  instant  relief  to  the 
schismogenic  strain,  even  though  the  schismogenesis  had 
previously  reached  the  limits  of  cultural  tolerance. 

3.  Similarly,  a  sudden  change  in  the  terms  of  a  sym- 
metrical rivalry  may  relieve  the  schismogenic  strain.  An 
example  (either  of  this  or  of  No.  2)  may  be  seen  in  latmul 
culture  in  what  I  have  termed  the  diversion  of  harshness  into 
buffoonery.  Here  we  have  a  sudden  shift  in  behaviour 
patterns,  prompted  perhaps  by  an  individual's  consciousness 
of  strain.  It  is  likely  that  this  shift  immediately  retards  or 
reverses  the  former  schismogenesis. 

4.  It  is  reasonably  certain  that  either  a  complementary 
or  a  symmetrical  schismogenesis  between  two  groups  can 
be  checked  by  factors  which  unite  the  two  groups  either  in 
loyalty  or  opposition  to  some  outside  element.  Such  an 
outside  element  may  be  either  a  symbolic  individual,  an 
enemy  people,  or  some  quite  impersonal  circumstance  which 
gives  satisfaction  or  misery  to  both  groups  alike.  Examples 
of  this  process  are  common  enough:  the  celebration  of  a 
Jubilee  is  useful  in  reducing  the  schismogenic  antipathy 
between  the  different  strata  of  society ;  and  an  outside  war  is 
well  known  as  a  device  for  reviving  a  nationalism  which 
suffers  from  schismogenic  splits.  In  latmul  culture  the  same 
mechanism  presumably  occurs.  It  is  probable  that  head- 
hunting warfare  against  outside  communities  contributed  in 
no  small  measure  to  control  the  schismogeneses  between  the 
sexes  and  between  the  initiatory  moieties;  and  it  is  possible 
that  nowadays  the  complementary  relationship  between  na- 
tives and  Europeans  has  to  some  extent  taken  over  analogous 
functions. 

But  we  may  note  that  wherever  the  outside  element  is  a 
person  or  a  community,  the  relationship  between  this  outside 
group  and  the  original  two  groups  is  itself  always  schismo- 
genic, and  that  this  second  schismogenesis  may  be  even  more 
difficult  to  control  than  the  first. 


and  Schismogenesis  195 

5.  The  process  mentioned  under  No.  4  presents  a  number 
of  problems,  the  answers  to  which  may  perhaps  reveal  other 
mechanisms  which  control  schismogenesis.  These  are  the 
problems  of  the  effect  of  an  external  schismogenesis  upon 
the  patterns  of  behaviour  zvithin  each  of  the  groups  involved 
in  it.  The  problem  may  be  stated  in  almost  mathematical 
form :  if  group  A  has  a  complementary  schismogenic  relation- 
ship with  group  B,  what  relation  can  we  observe  between  the 
patterns  of  behaviour  which  the  members  of  A  exhibit  in 
their  dealings  with  B  and  the  patterns  of  behaviour  which  they 
exhibit  in  their  dealings  with  one  another?  Similarly,  what 
are  the  patterns  of  behaviour  inside  group  B  among  indi- 
viduals who,  in  their  dealings  with  A,  exhibit  patterns  com- 
plementary to  those  exhibited  hy  A}  We  should  expect  that, 
if  the  schismogenic  relation  between  A  and  B  is  permitted 
to  proceed  to  a  point  at  which  marked  distortion  occurs  in 
the  personalities  concerned,  their  behaviour  inside  their  re- 
spective groups  should  be  markedly  affected  by  this  distortion ; 
and  since  the  direction  of  the  distortion  is  different  in  the  two 
groups,  we  should  find  that  in  each  group  there  is  developed 
a  special  ethos,  related  in  some  simple  way  to  the  terms  of 
the  schismogenic  contrast. 

When  the  problem  is  stated  in  this  way,  it  involves  two 
subsidiary  problems.  First,  that  of  the  mechanism  of  equi- 
libration of  hierarchies.  In  a  hierarchic  series,  we  may  see 
each  grade  as  involved  in  a  complementary  schismogenic 
relationship  with  the  grade  above,  and  in  another  similar 
relationship  with  the  grade  below.  We  may  suppose  that  the 
distortion  of  personality  caused  by  these  schismogeneses  is 
to  some  extent  relieved  by  the  fact  that  in  the  middle  groups 
of  the  series,  each  individual  is  prompted  to  exhibit  one  set 
of  patterns  of  behaviour  to  members  of  the  higher  groups 
and  a  set  of  complementary  patterns  to  members  of  the  lower 
groups.  But  this  compensation  is  not  permitted  to  the  highest 
and  the  lowest  members  of  the  series,  and  we  must  expect 
to  see  signs  of  unbalanced  schismogenesis  in  these  indi- 
viduals. 

The  second  problem  is  closely  related  to  that  of  the  hier- 
archies.   If  the  behaviour  between  individuals  within  each 


196  Ethological  Contrast^  Competition 

of  groups  A  and  5  is  a  function  of  the  behaviour  which  each 
group  exhibits  in  its  complementary  relationship  with  the 
other,  then  what  are  we  to  say  of  schismogeneses  which  occur, 
not  between  groups  of  individuals,  but  between  isolated  pairs 
of  individuals  ?  Are  we  to  suppose  that  compensations  occur 
within  the  personality  of  each?  And  is  it  possible  to  hope 
that,  when  we  know  how  the  behaviour  between  individuals 
is  influenced  by  the  schismogenesis  in  which  their  group  is 
involved,  an  understanding  of  this  process  will  give  us  a  jargon 
for  describing  the  compensations  which  occur  within  the 
individual  ?  ^ 

6.  Two  special  cases  of  the  control  of  schismogenesis  by 
outside  stimuli  may  be  mentioned :  {a)  the  possibility  that  if 
individuals  could  become  conscious  of  the  schismogenic  pro- 
cess in  which  they  are  involved,  they  might  react  to  this 
rather  than  continue  to  specialise  in  their  schismogenic 
patterns;  {h)  in  latmul  culture,  it  is  probable,  not  only  that 
the  schismogenic  relation  between  the  men  and  the  women 
contributes  to  controlling  the  symmetrical  schismogenesis  be- 
tween the  initiatory  moieties,  but  also  that  the  latter  schismo- 
genesis is  to  some  extent  controlled  by  the  orientation  of  the 
men's  attention  towards  the  secrets  of  initiation.  We  may  see 
here  a  complicated  example  involving  a  whole  series  of 
mechanisms  of  the  types  referred  to  under  Nos.  4  and  5. 

7.  It  is  possible  that  in  a  complementary  schismogenesis 
the  patterns  of  behaviour  between  the  two  groups  concerned 
may  be  such  that,  while  each  group  diverges  farther  and 
farther  from  the  other,  the  members  of  each  also  become  more 
and  more  dependent  upon  the  complementary  behaviour  of 
the  members  of  the  other  group,  so  that  at  some  point  in  the 
progress  of  the  schismogenesis  a  balance  will  be  reached 

^  In  the  case  of  schismogenesis  between  groups,  we  noted  that  a 
schismogenic  relationship  with  an  outside  enemy  might  prevent  schismo- 
genesis between  the  classes  within  a  nation,  but  we  also  noted  that  outside 
admiration  might  induce  a  schismogenic  split  within  the  personality  of  an 
individual.  This  is  probably  not  a  discrepancy  between  individual  and 
group  phenomena  but  an  indication  that  an  external  schismogenesis  will 
promote  an  internal  only  if  both  splits  are  in  the  same  terms.  A  war 
between  two  capitalist  states  may  reduce  the  schismogenesis  between 
Fascists  and  Communists  in  each ;  but  a  war  with  Russia,  especially  if 
phrased  as  a  "  war  against  Communism  ",  might  have  very  different  effects. 


and  Schismogenesis  197 

when  the  forces  of  mutual  dependence  are  equal  to  the 
schismogenic  tendency. 

8.  Lastly,  it  is  certain  that  a  schismogenesis,  besides  being 
controlled  by  other  schismogeneses  either  inside  or  outside 
the  groups  concerned,  may  also  be  counteracted  by  inverse 
progressive  changes  in  the  relationship  between  the  groups 
concerned.  These  processes  are,  like  schismogeneses,  cumula- 
tive results  of  each  individual's  reactions  to  the  reactions  of 
members  of  the  other  group,  but  the  inverse  process  differs 
from  schismogenesis  in  the  direction  of  the  change.  Instead 
of  leading  to  an  increase  in  mutual  hostility,  the  inverse 
process  leads  rather  in  the  direction  of  mutual  love. 

This  process  occurs  not  only  between  groups  but  also 
between  pairs  of  individuals ;  and  on  theoretical  grounds,  we 
must  expect  that  if  the  course  of  true  love  ever  ran  smooth, 
it  would  follow  an  exponential  curve. 


CHAPTER  XIV 

The  Expression  of  Ethos  in  Naven 

SEX  ETHOS   AND   NAVEN 

TH  E  most  important  generalisation  which  can  be  drawn 
from  the  study  of  latmul  ethos  is  that  in  this  society 
each  sex  has  its  own  consistent  ethos  which  contrasts 
with  that  of  the  opposite  sex.  Among  the  men,  whether  they 
are  sitting  and  talking  in  the  ceremonial  house,  initiating  a 
novice,  or  building  a  house — whatever  the  occasion — there  is 
the  same  emphasis  and  value  set  upon  pride,  self-assertion, 
harshness  and  spectacular  display.  This  emphasis  leads  again 
and  again  to  over-emphasis;  the  tendency  to  histrionic  be- 
haviour continually  diverts  the  harshness  into  irony,  which 
in  its  turn  degenerates  into  buffooning.  But  though  the 
behaviour  may  vary,  the  underlying  emotional  pattern  is 
uniform. 

Among  the  women  we  have  found  a  different  and  rather 
less  consistent  ethos.  Their  life  is  concerned  primarily  with 
the  necessary  routines  of  food-getting  and  child-rearing,  and 
their  attitudes  are  informed,  not  by  pride,  but  rather  by  a 
sense  of  "reality".  They  are  readily  co-operative,  and  their 
emotional  reactions  are  not  jerky  and  spectacular,  but  easy 
and  "natural".  On  special  occasions,  it  seems,  the  women 
exhibit  an  ethos  modelled  upon  that  of  the  men,  and  it  would 
appear  from  our  consideration  of  Preferred  Types  that  cer- 
tain women  are  admired  for  what  we  may  describe  as  latmul- 
masculine  characteristics. 

If  we  return  at  this  point  to  the  problems  presented  by 
the  naven  ceremonies,  we  see  these  problems  in  a  new  light. 
The  elements  of  exaggeration  in  the  wau's  behaviour  appear, 
not  as  isolated  oddities,  but  as  patterns  of  behaviour  which 
are  normal  and  ordinary  in  latmul  men.  This  answer  may 
seem  rather  uninteresting  but  it  involves  a  major  generali- 
sation about  cultural  behaviour,  and  in  science  every  step  is 
a  demonstration  of  consistency  within  a  given  sphere  of 


The  Expression  of  Ethos  in  Naven  199 

relevance.  We  might  perhaps  have  studied  this  consistency 
more  fully  but  the  answer  would  still  have  been  of  the  same 
type.  To  pursue  the  matter  further  we  should  be  compelled 
to  shift  to  some  other  scientific  discipline,  e.g.  to  the  study 
of  character  formation. 

In  the  case  of  the  women,  with  a  double  emphasis  running 
through  their  ethos,  their  naven  behaviour  can  be  completely 
classified  as  consistent  either  with  their  everyday  ethos  or 
with  their  special  occasional  pride.  All  the  behaviour  of  the 
mother  is  patterned  upon  submission  and  negative  self- 
feeling.  Her  action  in  lying  naked  with  the  other  women 
while  her  son  steps  over  them,  and  the  cliche  "that  so  small 
place  out  of  which  this  big  man  came",  are  perfectly  in 
keeping  with  the  everyday  ethos  of  latmul  women,  and 
constitute  a  very  simple  expression  of  her  vicarious  pride  in 
her  son.  Thus  the  problem  of  the  mother's  behaviour,  like 
that  of  the  waus  exaggerations,  may  now  be  referred  to  other 
scientific  disciplines. 

In  the  behaviour  of  the  transvestite  women,  the  father's 
sister  and  the  elder  brother's  wife,  we  may  see  an  expression 
of  the  occasional  pride  such  as  women  exhibit  on  the  rare 
occasions  when  they  perform  publicly  with  men  as  an  audience. 

Examination  of  latmul  ethos  has  accounted  for  the  tone 
of  behaviour  of  the  various  relatives  in  naven,  but  there  are 
many  details  which  cannot  be  thus  summarily  dismissed. 
Consider  the  wau:  his  buffooning  is  normal,  but  that  is  no 
reason  why  he  should  dress  as  a  woman  in  order  to  be  a 
buffoon,  and,  as  we  have  seen  above,  the  structural  premises 
within  the  culture,  whereby  the  wau  might  regard  himself 
as  the  laua's  wife,  are  still  not  a  dynamic  factor  which  would 
compel  either  the  community  or  the  wau  to  emphasise  this 
aspect  of  the  wau-laua  relationship.  We  have  still  to  find  some 
component  of  the  naven  situation  which  shall  act  in  a  dynamic 
way  to  induce  transvesticism. 

I  believe  that  we  may  find  an  answer  to  this  problem  if 
we  examine  the  incidence  of  transvesticism  in  European 
society.  In  the  naven,  the  phenomenon  is  not  due  to  abnormal 
hormones  nor  yet  to  the  psychological  or  cultural  malad- 
justment of  the  transvestites ;  and  therefore  in  looking  for 


200  The  Expression  of  Ethos  in  Naven 

analogous  phenomena  in  Europe  we  may  ignore  the  aberrant 
cases  and  should  examine  rather  the  contexts  in  which  some 
degree  of  transvesticism  is  culturally  normal. 

Let  us  consider  the  case  of  the  fashionable  horsewoman. 
Her  breeches  we  may  perhaps  regard  as  a  special  adaptation, 
and  she  will  say  that  her  bowler  hat  is  specially  designed  to 
protect  her  head  from  overhanging  trees:  but  what  of  her 
coat,  tailored  on  decidedly  masculine  lines?  She  wears 
feminine  evening  dress  at  the  hunt  ball,  and  her  everyday 
behaviour  is  that  of  a  culturally  normal  woman,  so  that  we 
cannot  explain  her  transvesticism  by  a  reference  to  her  glands 
or  abnormal  psychology. 

The  facts  of  the  matter  are  clear :  a  culturally  and  physically 
normal  woman  wears,  in  order  to  ride  a  horse,  a  costume 
unusual  for  her  sex  and  patterned  on  that  of  the  opposite 
sex ;  and  the  conclusion  from  these  facts  is  equally  obvious : 
since  the  woman  is  normal,  the  unusual  element  must  be 
introduced  by  the  act  of  riding  a  horse.  In  one  sense,  of 
course,  there  is  nothing  exceptional  in  a  woman's  riding — 
women  have  ridden  horses  for  hundreds  of  years  in  the 
history  of  our  culture.  But  if  we  compare  the  activity  of 
riding  a  horse  with  other  activities  which  our  culture  has 
decreed  suitable  and  proper  for  women,  we  see  at  once  that 
horse-riding,  which  demands  violent  activity  and  gives  a  great 
sense  of  physical  mastery,^  contrasts  sharply  with  the  great 
majority  of  situations  in  a  woman's  life. 

The  ethos  of  women  in  our  culture  has  been  built  up 
around  certain  types  of  situation  and  that  of  men  around  very 
different  situations.  The  result  is  that  women,  placed  by 
culture  in  a  situation  which  is  unusual  for  them  but  which  is 
usual  for  men,  have  contrived  a  transvestite  costume,  and  this 
costume  has  been  accepted  by  the  community  as  appropriate 
to  these  abnormal  situations. 

With  this  hint  of  the  sort  of  situation  in  which  trans- 

^  In  Freudian  phrasing  the  act  of  riding  a  horse  might  be  regarded 
as  sexually  symbolic.  The  difference  between  the  point  of  view  which 
I  advocate  and  that  of  the  Freudians  is  essentially  this :  that  I  regard 
such  sexual  symbols  as  noses,  flutes,  wagan,  etc.,  as  symbolic  of  sex  ethos, 
and  I  would  even  see  in  the  sexual  act  one  more  context  in  which  this 
ethos  is  expressed. 


The  Expression  of  Ethos  in  Naven  201 

vesticism  may  be  developed,  we  may  return  to  latmul  culture. 
First  let  us  consider  the  contexts  in  which  partial  trans- 
vesticism  occurs,  namely,  in  the  case  of  women  who  take  part 
in  spectacular  ceremonies  (cf.  Plate  XIX).  Their  position  is 
very  closely  analogous  to  that  of  the  horsewoman.  The  normal 
life  of  latmul  women  is  quiet  and  unostentatious,  while  that 
of  the  men  is  noisy  and  ostentatious.  When  women  take  part 
in  spectacular  ceremonial  they  are  doing  something  which  is 
foreign  to  the  norms  of  their  own  existence,  but  which  is 
normal  for  men — and  so  we  find  them  adopting  for  these 
special  occasions  bits  of  the  culture  of  the  men,  holding  them- 
selves like  men^  and  wearing  ornaments  which  are  normally 
only  worn  by  men. 

Looking  at  the  naven  ceremonies  in  the  light  of  this  theory 
we  can  recognise  in  the  naven  situation  conditions  which 
might  influence  either  sex  towards  transvesticism.  The  situa- 
tion may  be  summed  up  by  saying  that  a  child  has  accom- 
plished some  notable  feat  and  its  relatives  are  to  express,  in 
a  public  manner,  their  joy  in  this  event.  This  situation  is  one 
which  is  foreign  to  the  normal  settings  of  the  life  of  either 
sex.  The  men  by  their  unreal  spectacular  life  are  perfectly 
habituated  to  the  "  ordeal "  of  public  performance.  But  they 
are  not  accustomed  to  the  free  expression  of  vicarious  per- 
sonal emotion.  Anger  and  scorn  they  can  express  with  a  good 
deal  of  over-compensation,  and  joy  and  sorrow  they  can 
express  when  it  is  their  own  pride  which  is  enhanced  or 
abased ;  but  to  express  joy  in  the  achievements  of  another  is 
outside  the  norms  of  their  behaviour. 

In  the  case  of  the  women  the  position  is  reversed.  Their 
co-operative  life  has  made  them  capable  of  the  easy  expres- 
sion of  unselfish  joy  and  sorrow,  but  it  has  not  taught  them 
to  assume  a  public  spectacular  role. 

Thus  the  naven  situation  contains  two  components,  the 

^  In  theatrical  representations,  humorous  journalism  and  the  like, 
there  is  a  common  belief  that  the  postures,  gestures,  tones  of  the  voice, 
etc.,  of  the  horsewoman  are  to  some  extent  modelled  on  those  of  men, 
and  we  might  see  in  this  an  analogy  with  the  proud  gestures  of  the  trans- 
vestite  and  semi-transvestite  latmul  women.  I  am  uncertain,  however, 
to  what  extent  these  postures,  etc.,  of  the  horsewoman  occur  in  real  life, 
and  whether  they  are  not  perhaps  imaginary. 


202  The  Expression  of  Ethos  in  Naven 

element  of  public  display  and  the  element  of  vicarious  per- 
sonal emotion ;  and  each  sex,  when  it  is  placed  by  culture  in 
this  situation,  is  faced  by  one  component  which  is  easily 
acceptable,  while  the  other  component  is  embarrassing  and 
smacks  rather  of  situations  normal  to  the  life  of  the  opposite 
sex.  This  embarrassment  we  may,  I  think,  regard  as  a  dy- 
namic^ force  which  pushes  the  individual  towards  trans- 
vesticism — and  to  a  transvesticism  which  the  community  has 
been  able  to  accept  and  which  in  course  of  time  has  become 
a  cultural  norm. 

Thus  the  contrasting  ethos  of  the  two  sexes  may  be  sup- 
posed to  play  and  to  have  played  in  the  past  a  very  real  part 
in  the  shaping  of  the  naven  ceremonies.  It  has  provided  the 
little  push  which  has  led  the  culture  to  follow  its  structural 
premises  to  the  extremes  which  I  have  described.  When  the 
women  take  part  in  spectacular  ceremonial  other  than  naven, 
the  structural  premises  which  might  justify  complete  trans- 
vesticism are  lacking,  and  the  women  content  themselves  with 
wearing  only  a  few  masculine  ornaments. 

Lastly  we  may  consider  the  adoption  of  widow's  weeds  by 
the  wau  and  the  wearing  by  the  women  of  the  best  masculine 
ornaments  obtainable.  The  former  is  no  doubt  a  buffooning 
expression  of  the  men's  distaste  for  the  women's  ethos.  We 
have  seen  that  the  context  of  mourning  is  one  in  which  the 
differing  ethos  of  the  two  sexes  contrasts  most  strongly  and 
most  uncomfortably,  and  the  wearing  of  a  widow's  weeds  by 
the  wau  is  clearly  on  all  fours  with  the  men's  trick  of  carica- 
turing the  dirging  of  the  solitary  widow  as  she  paddles  her 
canoe  to  her  garden.  In  shaming  himself  he  is,  incidentally, 
expressing  his  contempt  for  the  whole  ethos  of  those  who 
express  grief  so  easily. 

The  women  on  the  other  hand  have  no  discernible  con- 
tempt for  the  proud  male  ethos.  It  is  the  ethos  appropriate 
to  spectacular  display,  and  in  the  naven  they  adopt  as  much 
of  that  ethos  as  possible — and  even  exaggerate  it,  gaily 
scraping  the  lime  sticks  in  their  husbands'  gourds  till  the 
serrations  are  quite  worn  away.    Here  it  would  seem  that 

^  Cf.  footnote,  p.  121. 


The  Expression  of  Ethos  in  Naven  203 

their  joy  in  wearing  masculine  ornaments  and  carrying  on  in 
the  swaggering  ways  of  men  has  somewhat  distracted  them 
from  the  business  in  hand — that  of  celebrating  the  achieve- 
ment of  a  small  child.  Apart  from  the  one  incident  in  which 
the  women  lie  down  naked  while  the  hero  steps  over  them, 
the  naven  behaviour  of  the  women  is  actually  as  irrelevant 
as  that  of  the  men.  Thus  the  presence  of  contrasting  ethos 
in  the  two  sexes  has  almost  completely  diverted  the 
naven  ceremonial  from  simple  reference  to  its  ostensible 
object. 

Nevertheless,  since  naven  behaviour  is  the  conventional 
way  in  which  a  wau  congratulates  his  laua  upon  any  achieve- 
ment, there  is  no  doubt  that  this  behaviour,  distorted  and 
irrelevant  as  it  may  seem  to  us,  is  yet  understood  by  the  laua 
as  a  form  of  congratulation. 


KINSHIP  MOTIVATION  AND  NAVEN 

I  have  shown  above  how  the  naven  ceremonial  is  an  expres- 
sion of  sex  ethos,  that  is  to  say,  I  have  shown  the  motives 
which  we  may  attribute  to  the  relatives  concerned  in  naven 
qua  members  of  one  or  the  other  sex.  And  this  attribution 
of  motive  was  an  extrapolation  from  general  conclusions 
which  we  drew  from  the  ethos  of  latmul  culture.  We  have 
now  to  consider  the  motivation  of  the  various  relatives  qua 
men  and  women  in  particular  kinship  positions.  Here  again 
we  shall  use  our  general  ethological  conclusions  to  guide  us 
in  selecting  the  motives  which  we  attribute  to  the  various 
relatives.  But  instead  of  dealing  with  the  relatives  in  groups 
according  to  their  sex,  we  must  examine  the  kinship  motiva- 
tion of  each  type  of  relative  separately : 

wau  (mother's  brother).  In  the  case  of  the  wau,  the 
symbolism  of  his  behaviour  is  in  a  structural  sense  clearly 
interpretable  in  terms  of  his  position  as  "  mother  "  and  "wife  " 
of  the  laua.  But  symbolism  has  always  more  than  a  struc- 
tural side,  and  we  should  see  the  symbolic  acts — the  rubbing 
of  the  male  laua's  shin  with  the  buttocks,  and  the  giving 


204  The  Expression  of  Ethos  in  Naven 

birth  to  the  female  laua — as  acts  informed  with  emotion 
and  inarticulate^  reference,  as  well  as  with  "logic". 
We  have  therefore  to  consider  what  motivation,  conscious 
or  otherwise,  real  or  culturally  presumed,  leads  the  wau 
to  stress  these  particular  aspects  of  his  relationship  with  his 
laua. 

The  undoubted  fact  that  any  emotion  and  inarticulate 
motivation  which  we  may  impute  to  the  classificatory  wau 
will  be  factitious  need  not  cause  us  any  dismay;  indeed,  it 
would  be  difficult  to  explain  any  ritual  without  invoking  as 
its  mainstay  the  very  real  force  of  trumped-up  or  culturally 
presumed  emotion. 

In  the  chapter  dealing  with  the  sociology  of  naven  we  have 
seen  that  the  classificatory  wau  is  attempting  to  bolster  up 
the  allegiance  between  himself  and  his  laua.  He  is  trying  to 
bring  this  classificatory  tie  into  such  prominence  that  it  shall 
have  the  binding  power  of  a  tie  between  a  boy  and  his  own 
mother's  brother — perhaps  even  the  binding  power  of  the 
tie  with  own  mother.  Symbolically,  he  does  this  by  acting 
out  two  sides,  the  maternal  and  the  wifely,  of  the  own  mother's 
brother's  relationship  with  the  laua.  But  it  is  questionable 
to  what  extent  real  wifely  or  maternal  emotion  can  be  at- 
tributed even  to  the  own  wau\  so  that  the  performance  of 
the  classificatory  zvau  contains,  in  a  sense,  two  degrees  of 
falsity.  He  is  acting  the  part  of  an  own  wau  who  is  acting 
the  parts  of  mother  and  wife. 

In  spite  of  this,  it  is  still  worth  while  to  consider  the  motiva- 
tion of  the  performances.  The  fact  that  the  behaviour  is  acted 
rather  than  spontaneous  will  not,  of  course,  render  the  motiva- 
tions which  it  portrays  any  the  less  unconscious.  An  actor 
playing  Hamlet  may  behave  as  if  he  himself  were  driven  by 
Hamlet's  emotions,  but  the  actor  may  remain  to  a  great 
extent  inarticulate  as  to  the  nature  of  the  drives  which  he 
is  expressing. 

^  In  the  present  context,  I  have  deliberately  written  "inarticulate"  in 
preference  to  the  more  fashionable  "unconscious".  It  is  to  me  not 
perfectly  clear  that  there  is  any  diflTerence  in  meaning  between  these  two 
terms,  but  the  use  of  the  second  would  imply  adherence  to  a  theory  of 
the  stratification  of  mind  with  the  validity  of  which  we  are  not  here 
concerned. 


The  Expression  of  Ethos  in  Naven  205 

We  have  to  consider,  then,  the  psychology  of  two^  persons : 

1.  The  real  motivation  of  the  classificatory  wau.  Under 
this  head  will  fall  the  series  of  emotional  factors  which  have 
already  been  sufficiently  indicated:  his  pleasure  in  his  own 
buffooning  performance ;  his  scorn  of  the  female  ethos ;  his 
awkwardness  in  expressing  vicarious  pride;  his  desire  for 
the  allegiance  of  the  laua,  etc. 

2.  The  inarticulately  presumed  motivations  of  the  own 
wau.  Under  this  head  we  have  to  consider  possible  drives 
which  might  lead  a  wau  to  stress  the  maternal  and  wifely 
sides  of  his  relationship  with  his  laua. 

Here  we  are  on  very  treacherous  ground,  since  it  is  singu- 
larly easy  to  construct  hypotheses  about  inarticulate  motiva- 
tion, but  exceedingly  difficult  to  test  them,  since  no  two  such 
hypotheses  are  ever  mutually  exclusive,  and  the  psycho- 
analysts will  tell  us  to  regard  every  native  statement  as 
meaning  either  what  it  says,  or  the  opposite.  I  shall,  however, 
put  forward  a  series  of  possible  hypotheses  in  order  to  indi- 
cate what  sort  of  answer  may  be  expected  in  the  present  case, 
but  I  would  not  stress  any  one  hypothesis  as  more  probable 
than  others  which  might  be  constructed. 

In  this  connection  there  are  three  myths  which  are  probably 
relevant.  These  are  all  stories  in  which  a  wau  kills  his  laua. 
In  each  case  the  murder  is  followed  by  a  more  or  less  com- 
plete disintegration  of  the  community,  and  the  myth  be- 
comes a  history  of  colonisation — of  the  scattering  of  groups 
from  the  disintegrated  centre. 

These  myths  may  be  (i)  a  reference  to  the  importance  of 
matrilineal  ties  in  the  integration  of  the  community,  or  (2)  the 
expression  of  an  underlying  opposition  or  slight  hostility 
between  wau  and  laua.  These  interpretations  are  not  mutually 
exclusive.  Rather,  I  believe  that  we  should  see  the  two  inter- 
pretations as  mutually  supplementary.  On  the  one  hand,  the 
sociological  importance  of  the  tie  may  well  lead  to  an  increased 
uneasiness  about  the  opposition  which  is  inherent  in  it,  while 
on  the  other  hand,  if  there  were  no  opposition  between  wau 
and  laua,  there  would  be  no  need  to  stress  the  sociological 

^  In  the  rarer  cases  in  which  own  wau  performs  naven,  the  matter  is,  of 
course,  simpler,  and  the  second  set  of  emotions  proportionately  more  real. 


2o6  The  Expression  of  Ethos  in  Naven 

importance  of  allegiance  between  them.  In  general,  the 
myths  may  be  taken  as  an  indication  of  some  opposition 
between  wau  and  laua  and  some  uneasiness  about  that  opposi- 
tion. In  fact,  it  would  appear  that  the  wau-laua  relationship 
is  ambivalent. 

It  is  in  such  ambivalent  feelings  that  we  may  expect  to 
find  the  motives  for  the  zvau's  symbolic  behaviour  and  we 
must  therefore  examine  the  wau-laua  relationship  to  find 
what  aspects  of  it  may  be  regarded  as  the  cause  of  these 
mixed  feelings. 

We  have  not  far  to  seek  for  the  basis  of  the  wau's  friendly 
feelings  for  his  laua,  since  the  two  are  united  by  ties  of 
allegiance  and  conmion  interest.  Following  the  lines  of  our 
structural  analysis,  we  may  attribute  the  zoau's  friendly 
feelings  to  any  and  all  of  the  four  identifications  which  define 
his  position.  But  the  basis  for  the  trace  of  hostility  is  more 
obscure.  In  our  structural  analysis  we  found  that  the  wau- 
laua  relationship  contains  little  or  nothing  which  might  be 
regarded  as  sui  generis,  and  that  all  its  details  could  be  classi- 
fied satisfactorily  as  due  to  the  various  identifications  of  the 
wau  as  (a)  mother,  (b)  wife,  (c)  wife's  brother,  and  (d)  father 
of  his  laua. 

It  is  therefore  reasonable  in  our  search  for  the  basis  of  the 
wau's  hostihty  to  suppose  that  this  arises  from  one  or  other 
of  these  aspects  of  the  relationship.^  Of  these  aspects,  the 
last,  that  due  to  the  waus  identification  with  the  father,  might 
perhaps  give  a  possible  phrasing  of  the  waus  hostility  as  an 
expression  of  the  ambivalent  situation  between  father  and 
son.  But  I  think  that  this  solution  must  be  dismissed,  for 
two  reasons: 

1.  This  solution  is  incompatible  with  the  fact  that,  not 
only  the  wau,  but  also  the  wife's  brother  sometimes  rubs  his 
buttocks  on  his  sister's  husband's  leg,  when  the  latter  marries 
his  sister. 

2.  The  identification  of  wau  with  father  is  only  a  very 
minor  motif  in  the  patterning  of  the  wau-laua  relationship. 

^  It  is,  of  course,  possible  that  the  luau's  "hostility"  may  be  due  in 
part  to  interaction  between  two  or  more  of  these  aspects  of  his  position. 
This  complicated  type  of  explanation  I  have  ignored  in  favour  of  the 
simpler  solution  based  upon  the  brother-in-law  relationship. 


The  Expression  of  Ethos  in  Naven  207 

Of  the  three  remaining  aspects  of  the  wau-laua  relation- 
ship, two,  the  maternal  and  the  wifely,  may  likewise  be 
excluded  on  various  grounds : 

1.  These  are  the  aspects  which  are  actually  stressed  in 
the  naven  and,  though  it  is  conceivably  possible  that  the  wau 
might  stress  those  very  aspects  which  were  the  cause  of  his 
trace  of  hostility,  it  is  much  more  likely  that  he  gets  relief 
from  his  uneasiness  by  stressing  the  other  aspects. 

2.  It  is  exceedingly  difficult  and  perhaps  impossible,  if  we 
attribute  the  basis  of  the  opposition  to  either  of  these  aspects, 
to  construct  any  theory  which  will  cover  the  waw's  naven 
behaviour  to  lauas,  of  both  sexes,  and  the  naven  behaviour 
of  the  wife's  brother, 

3 .  The  mother's  relationship  to  her  child  is  not  ambivalent. 
We  are  therefore  driven  to  conclude  that  the  zvau's  trace 

of  hostility  is  due  to  his  position  as  a  brother-in-law  of  his 
laua.  And  on  this  theory  we  may  see,  in  his  stressing  of  the 
"mother"  and  "wife"  relationships,  a  denial  of  his  position 
as  wife's  brother. 

This  theory  receives  support  from  examination  of  the 
brother-in-law  relationship,  since  we  can  clearly  detect  ele- 
ments of  opposition^  in  that  relationship  (cf.  p.  78).  But  the 
position  is  still  not  perfectly  clear.  The  familiar  processes  of 
compensation  and  displacement  cannot  be  appealed  to  in 
the  present  case.  Typically,  the  wau  having  ambivalent  feelings 
towards  his  sister's  husband  might  either  exaggerate  his 
friendliness  in  his  dealings  with  that  relative,  or  he  might 
vent  his  hostility  upon  the  laua  who  would  then  be  a  substi- 
tute for  the  sister's  husband.  In  practice,  he  does  neither 
of  these  things,  but  exaggerates  his  friendliness  towards  his 
laua  rather  than  his  hostility.  Further,  there  is  no  reason 
why  the  hostility  felt  for  the  sister's  husband  should  be  dis- 
placed on  to  the  laua.  We  have  seen  that  in  general  this 

^  It  should  be  possible  to  demonstrate  that  tatvontu's  (wife's  brother's) 
hostility  is  consistent  with  the  male  ethos  of  the  culture,  and  I  think 
that  the  clue  to  this  problem  perhaps  lies  in  a  pride  which  the  men  occa- 
sionally exhibit,  a  pride  in  the  possession  of  women-folk.  Cf.  also  the 
cliche  "She  is  a  fine  woman"  as  a  justification  for  endogamy  (p.  91). 
Unfortunately  I  have  not  the  detailed  material  necessary  for  a  comparison 
of  attitudes  of  men  towards  their  sisters  with  the  general  male  ethos. 


2o8  The  Expression  of  Ethos  in  Naven 

hostility  is  openly  handled  and  acknowledged.  The  brothers- 
in-law  may  either  joke  about  their  mutual  obligations ;  or,  if 
the  wife's  brother's  resentment  grows  strong,  he  is  expected 
to  have  recourse  to  sorcery.  It  is  true  that  the  quarrels  of 
brothers-in-law  are  phrased  in  terms  of  indebtedness  and 
bride  price  in  a  context  in  which  a  Freudian  might  suspect 
sexual  jealousy  to  be  the  underlying  motive,  but  still  the 
hostile  feelings  of  the  wife's  brother  are  in  no  sense  denied 
expression.  There  is  nothing  in  all  this  which  would  lead 
him  to  seek  a  substitute  for  his  sister's  husband. 

It  is,  however,  possible  to  see  the  wau  as  a  man  guilty  of 
overt  hostility  to  his  sister's  marriage,  who  therefore  makes 
exaggerated  amends  to  her  child,  and  a  tenable  hypothesis 
can  be  constructed  upon  this  basis.  But,  equally,  we  may 
see  the  wau,  not  as  a  guilty  man  making  amends,  but  as  an 
innocent  man  protesting  his  innocence.  He  may  have  op- 
posed his  sister's  marriage  and  have  harboured  ill-feeling 
for  her  husband.  In  any  case  it  is  culturally  assumed  that 
some  opposition  exists  between  brothers-in-law.  But  he  has 
no  quarrel  with  the  offspring  of  that  marriage,  who  are  indeed 
connected  with  him  by  blood,^  interest  and  allegiance.  The 
offspring,  however,  are  also  connected  no  less  intimately  with 
the  father,  and  it  is  this  fact  which  might  constitute  an 
imputation  that  the  wau  is  perhaps  hostile  towards  them.  It 
suggests  that  the  zvau's  hostility,  previously  centred  on  the 
sister's  husband,  might  perhaps  spread  to  include  the  laua. 
This  imputation  the  zvau  is  perhaps  denying  when  he  stresses 
the  fact  that  he  is  a  mother  and  a  wife. 

Similar  phenomena  occur  in  our  own  culture,  and  it  is  not 
uncommon  to  see  uncles  and  aunts  who  have  consistently 
opposed  the  parents'  marriage  later  falling  over  themselves 
in  their  efforts  to  make  friends  with  their  nephews  and  nieces. 
Their  position  is  analogous  to  that  of  the  latmul  wau,  and 
the  chief  difference  between  them  is  that  the  latter  falls  over, 
not  metaphorically,  but  really. 

Whether  we  see  these  agnatic  relatives  as  innocent  persons 

^  In  the  native  theory  of  conception,  blood  and  flesh  are  believed  to 
be  products  of  the  mother,  while  the  bones  of  the  child  are  contributed 
by  the  father. 


The  Expression  of  Ethos  in  Naven  209 

protesting  their  innocence  or  as  guilty  persons  making  amends, 
will  depend — in  the  absence  of  additional  facts — upon  the 
view  which  we  take  of  psycho-analytic  theory.  We  may  even, 
according  to  our  taste,  combine  the  two  hypotheses  and  say 
that  the  wau,  guilty  of  overt  hostility  towards  his  sister's 
husband,  therefore  protests  his  innocence  with  the  greater 
emphasis  when  a  similar  offence  is  imputed  to  him  by  the 
circumstances  of  his  relationship  with  his  laua,  about  whom 
he  really  feels  innocent.  A  whole  series  of  varying  phrasings 
of  this  sort  may  be  devised,  and  each  of  them  will  perhaps 
be  untrue,  because  it  will  err  on  the  side  of  being  too  definite ; 
while  any  attempt  to  choose  between  them  will  be  based 
upon  the  fallacy  that  the  less  articulate  levels  of  the  mind 
are  as  discriminating  and  as  precise  as  the  scientist.^ 

In  any  case,  we  may  add  to  our  list  of  factors  affecting 
the  zvau's  position  the  opposition  which  is  present  in  the 
brother-in-law  relationship  and  its  repercussions  upon  that 
between  wau  and  laua^  without  defining  the  exact  mechanism 
of  these  repercussions.  Further,  it  may  be  noted  that  these 
additional  factors  may  be  regarded  as  to  some  extent  dynamic 
— as  pushing  the  wau  towards  transvesticism  along  the  lines 
laid  down  in  the  structure  of  the  kinship  system. 

Finally,  one  other  possible  motive  for  the  wam's  behaviour 
in  naven  may  be  mentioned.  We  have  seen  in  our  considera- 
tion of  the  boy's  relationship  to  his  father's  and  his  mother's 
clans  that,  while  in  matters  of  economics  the  boy  is  grouped 
with  his  father,  his  feats  are  regarded  as  the  achievements  of 
his  mother's  clan  (cf.  p.  48).  In  view  of  these  premises,  we 
may  suppose  that  the  naven  behaviour  of  the  wau  is  an  act 
which  symbolically  lays  claim  to  the  laua's  achievement,  and 
the  value  which  this  culture  sets  upon  pride  and  achievement 
no  doubt  here  plays  a  part.  But  such  a  claim  can  only  be 
vicarious,  and  here  the  male  sex  ethos  is  again  relevant  and 

^  It  is,  however,  possible  that  a  more  detailed  study  of  the  standardised 
attitudes  and  behaviour  of  the  men,  when  though  innocent  they  are 
accused  and  again  when,  guilty,  they  are  making  amends,  might  make 
possible  a  more  definite  statement  of  the  zvati's  motives.  To  some  extent, 
at  least,  the  vagueness  of  the  statement  here  offered  is  the  inevitable  result 
of  attempting  an  analysis  of  motives  with  an  insufficient  knowledge  of 
the  ethology  of  the  culture. 


2IO  The  Expression  of  Ethos  in  Naven 

indicates  that  in  this  act  the  wau  will  be  playing  the  role  of 
a  woman.  Thus  sex  ethos  and  kinship  position  work  together, 
and  each  is  expressed  in  the  waiis  behaviour. 

tawontu  (wife's  brother).  From  what  has  been  said  of  the 
wau  in  naven,  it  might  appear  that  the  tawontu's  behaviour, 
in  rubbing  his  buttocks  on  his  lando's  shin  on  the  occasion 
of  the  latter's  marriage,  is  in  some  sense  the  primary  form 
of  naven.  We  are,  however,  not  engaged  in  a  study  of  the 
history  of  naven,  and  it  is  certain  that  at  the  present  time  this 
particular  gesture  is  regarded  as  especially  characteristic  of 
the  wau. 

The  factors  which  affect  the  two  relatives  are  alike  in  many 
respects,  but  the  behaviour  of  tawontu  differs  from  that  of 
wau  in  certain  significant  details:  (i)  The  tawontu  performs 
only  that  ritual  which  stresses  his  position  as  "wife"  and 
omits  that  which  would  imply  that  he  is  a  mother — an 
omission  which  is  no  doubt  correlated  with  his  structural 
position.  (2)  The  tawontu,  so  far  as  I  know,  only  performs 
his  gesture  on  the  occasion  of  the  marriage  of  his  sister. 
(3)  It  is,  I  believe,  only  the  own  wife's  brother  who  performs 
in  this  way,  whereas  in  the  case  of  wau,  it  is  usually  the  classi- 
ficatory  relative  who  performs  naven. 

The  second  and  third  of  these  points  of  difference  are 
probably  connected  with  the  fact  that  in  general  the  tawontu 
is  not  constrained  to  hide  or  deny  the  ambivalence  of  his 
feelings.  But  on  the  occasion  of  his  sister's  marriage — a 
marriage  which  he  himself  has  perhaps  arranged  (cf.  p.  79) — 
it  is  understandable  that  he  may  feel  constrained  to  deny 
the  existence  of  any  trace  of  hostility  ^  towards  her  husband ; 
and  this  denial  he  expresses  by  emphasising  the  fact  that 
he  is  not  an  affinal  relative,  but  a  partner  in  the  marriage, 
a  "wife-man ".2 

nyai' ,  nyamun  and  tshuamho  (father,  elder  brother  and 
younger  brother).  The  only  other  male  relatives  whose  posi- 
tion we  have  to  consider  are  the  father  and  brothers  of  the 

^  As  in  the  case  of  wau,  various  phrasings  of  the  tawontu  s  motives 
may  be  substituted  for  that  here  given. 

'  Literal  translation  of  the  term  tawontu  (tagwa-ndo  =  wiie-man). 


The  Expression  of  Ethos  in  Naven  211 

hero.  All  these  relatives  may  take  a  passive  part  in  naven; 
they  may  be  beaten  by  their  tshaishi  (elder  brothers'  wives) 
and  the  father  may  help  the  hero  to  find  valuables  to  present 
to  his  wau.  But,  apart  from  being  thus  secondarily  involved, 
neither  father  nor  brothers  take  any  active  part  in  naven. 

Since,  however,  his  father  and  brothers  are  the  boy's 
nearest  male  relatives,  we  might  reasonably  expect  that  they 
too  would  feel  impelled  to  celebrate  his  achievements;  it 
therefore  becomes  necessary  to  account  for  their  inactivity. 

We  may  first  consider  the  conditions — sociological,  eco- 
nomic, structural  and  emotional — which  are  probably  con- 
nected with  the  father's  inactivity : 

1 .  There  is  no  increase  in  the  integration  of  the  community 
to  be  gained  by  a  further  stressing  of  the  patrilineal  links. 
In  such  a  society  as  that  of  the  latmul,  the  solidarity  within 
any  group  depends  to  a  very  great  extent  upon  opposition 
between  that  group  and  others  outside  it:  at  the  same  time, 
it  is  probable  that  the  greater  the  solidarity  within  a  group 
and  the  greater  its  strength,  the  more  likely  it  is  to  oppose 
other  groups  or  to  ignore  its  obligations  to  them.  From  this 
it  follows  that  any  further  strengthening  of  the  patrilineal 
ties  may  lead  to  a  weakening  of  those  matrilineal  and  affinal 
ties  upon  which,  as  we  have  seen,  the  integration  of  the  com- 
munity depends. 

2.  On  the  economic  side,  there  is  no  occasion  for  presenta- 
tion of  valuables  between  father  and  son,  because  the  respec- 
tive property  rights  of  these  two  relatives  are  very  much  less 
differentiated  than  those  of  wau  and  laua.  Father  and  son 
do  not  hold  their  property  on  a  communal  basis  and  there 
is,  as  we  have  seen,  a  strong  feeling  against  the  father  accepting 
food  from  the  son:  but  there  is  a  general  assumption  that 
the  son  will  inherit  or  is  already  inheriting  from  the  father, 
and  that  the  father  will  do  what  he  can  to  provide  his  son 
with  valuables  for  bride  price,  etc.  Such  gifts  from  father  to 
son  are  not  in  any  way  comparable  with  the  ceremonial 
presentations^  which  take  place  between  wau  and  laua. 

^  It  would  seem  that  in  latmul  culture  ceremonial  giving  does  not 
occur  in  purely  complementary  relationships.  It  is  perhaps  a  device  for 
stressing  the  complementary  aspects  of  mixed  relationships.  (Cf.  p.  270.) 


212  The  Expression  of  Ethos  in  Naven 

Similar  considerations  apply  to  the  relationship  between 
brothers.  They  do  not  hold  their  property  on  a  communal 
basis,  but  the  differentiation  of  their  respective  rights  is  still 
so  vague  that  they  are  culturally  expected  to  quarrel  over 
their  patrimony. 

3 .  There  are  no  elements  in  the  structural  position  whereby 
either  father  or  brothers  might  divert  their  celebration  of  the 
boy's  achievement  into  some  form  of  buffoonery,  by  acting 
the  part  of  some  other  relative.  If  they  were  to  take  any  part 
in  the  naven  they  would  be  compelled  (like  the  mother)  to 
stress  their  own  relationships  with  the  hero. 

4.  This  last  consideration  brings  us  to  a  very  real  positive 
bar  on  any  naven  activity  on  the  part  of  the  father.  His 
relationship  with  his  son  is,  as  we  have  seen,  ambivalent; 
but  the  ambivalence  is  not  of  a  sort  which  can  be  relieved 
by  stressing  either  of  its  components.  The  father  is,  on  the 
one  hand,  a  disciplinary  authority  and,  on  the  other,  he  is  a 
relative  who  is  expected  to  give  way  before  his  son's  advance- 
ment. The  excessive  stressing  of  either  of  these  aspects  of 
the  paternal  relationship  leads  at  once  to  the  embarrassment 
of  both  parties,  and  we  have  seen  that  in  daily  life  father  and 
son  pattern  their  behaviour  on  a  rigid  middle  course,  avoiding 
even  any  display  of  mutual  intimacy.  In  such  a  situation  it 
is  clear  that  the  father  could  not  in  the  naven  stress  or 
dramatise  either  aspect  of  his  relationship  with  his  son. 

It  would  seem  that  even  the  passive  role  played  by  the 
father,  in  being  beaten  by  his  tshaishi,  is  to  some  extent 
incompatible  with  his  kinship  position.  I  was  told  (but  I  did 
not  observe  this)  that  the  own  father  would  be  only  slightly 
beaten,  but  the  tshambwi  nyai'  (father's  younger  brother) 
would  be  very  much  more  severely  beaten. 

We  may  note  the  contrast  between  the  father's  position 
and  that  of  the  wau.  The  latter's  kinship  position  is  defined 
in  terms  of  his  various  identifications ;  one  of  these  identifica- 
tions {qua  wife's  brother)  is  embarrassing,  and  he  escapes 
from  this  by  emphasising  the  others.  But  the  father  has  a 
unitary  relationship  with  his  son  which  is  not  subdivisible 
in  this  way.  Although  his  feelings  towards  his  son  are  cul- 
turally presumed  to  be  ambivalent,  he  cannot  dodge  the  whole 


The  Expression  of  Ethos  in  Naven  213 

issue  by  stressing  some  other  side  of  his  position,  and  if  he 
stressed  either  half  of  his  ambivalent  position  he  would  be 
embarrassed. 

In  the  case  of  elder  and  younger  brothers,  the  positive 
bar  is  less  clear:  but  the  relationship  between  brothers  is 
certainly  felt  to  be  analogous  to  that  between  father  and  son, 
though  the  issues  between  them  are  less  dramatically  and  less 
rigidly  defined.  On  the  one  hand,  brothers  are  expected  to 
be  allies  as  over  against  outsiders,  but  between  themselves 
they  are  culturally  expected  to  quarrel  over  the  patrimony, 
and  the  elder  brother  has  some  authority  over  the  younger. 
I  was  told  (but  never  observed  this)  that  there  is  some 
avoidance  between  brothers:  they  would  not  walk  about 
together  much;  they  would  go  together  in  order  to  do  some- 
thing, but  would  not  do  so  without  a  definite  purpose.  Thus 
their  relationship  contains  the  germ  of  the  same  type  of 
ambivalence  as  that  which  exists  between  father  and  son.^ 

Turning  now  to  the  naven  behaviour  of  the  women  we  find 
a  very  considerable  difference  between  their  motivation  and 
that  of  the  men.  It  is  possible  that  this  impression  is  due 
to  the  inadequacy  of  my  information,  but  I  believe  that  the 
contrast  is  a  real  one. 

In  the  case  of  the  wau  we  have  found  that  his  behaviour 
is  an  expression  of  the  following  factors : 

1.  His  structural  position  in  the  kinship  system  and  the 
various  identifications  which  define  it. 

2.  The  male  sex  ethos  of  latmul  culture. 

3.  His  need  for  the  allegiance  of  his  laua. 

4.  The  ambivalence  implied  in  his  kinship  position. 

5.  Economic  considerations. 

But  the  women,  so  far  as  I  know,  are  not  influenced  in  their 
naven  by  any  motives  of  the  last  three  types.  Rather,  I  believe 
that  we  should  see  the  women's  naven  as  almost  solely  the 
expression  of  their  ethos  and  of  the  structural  identifications 
implied  in  their  kinship  status.   I  am  inclined  to  see  the  naven 

^  As  evidence  that  this  relationship  is  really  felt  to  be  analogous  to 
that  between  father  and  son  cf.  the  staggering  of  brothers  (p.  246)  and 
the  use  of  the  refrain  " nyaV  nya!  nyamun  a!"  in  name  songs  (footnote, 
P-  39)- 


214  The  Expression  of  Ethos  in  Naven 

of  the  zcau  as  in  some  sense  primary  and  as  seriously  moti- 
vated by  desire  for  allegiance,  economic  gain,  etc.,  while 
I  would  see  the  naven  of  the  women  as  an  amusement,  a  gay 
occasion  when  they  embroider  upon  structural  premises 
analogous  to  those  followed  by  the  wau  and  when  they  enjoy 
the  special  privilege  of  wearing  men's  attire. 

In  the  case  of  the  wau,  we  observed  that  only  one  or  two 
waiis  usually  take  part,  and  that  these  were  especially  the 
classificatory  relatives.  But  there  is  no  such  rule  in  the  case 
of  the  women.  When  the  children  came  back  from  sago- 
making  and  a  naven  was  celebrated  for  them  in  Mindimbit, 
it  appeared  as  if  all  the  women  had  gone  mad.  Every  woman 
who  could  regard  herself  as  appropriately  related  to  the 
children  made  a  show  of  naven  behaviour. 

Moreover,  in  the  naven  for  the  successful  homicide  I  was 
told  that  all  the  women,  with  the  exception  of  the  hero's 
wife  and  own  sister,  lie  down  on  the  ground  for  the  hero 
to  stride  across  them.  In  this  detail  we  see  a  pattern  of 
behaviour,  which  is  perhaps  primarily  characteristic  of  the 
mother,  being  adopted  by  the  other  women  not  according 
to  the  details  of  their  kinship  status,  but  simply  in  virtue  of 
their  sex.  While  the  naven  of  the  men  is  in  part  controlled 
and  motivated  by  advantages  which  the  exploitation  of  their 
kinship  position  may  bring,  that  of  the  women  is  apparently 
almost  devoid  of  such  motivation. 

Correspondingly,  on  the  sociological  side  I  believe  that 
we  should  see  the  function  of  the  women's  naven  as  much 
more  diffuse  than  that  of  the  wau's  naven.  While  the  latter 
strengthens  the  solidarity  of  the  community  by  emphasising 
certain  specific  afiinal  linkages,  the  former  seems  rather  to 
spread  a  more  diffuse  euphoria  throughout  the  community. 

One  other  point  which  the  naven  behaviour  of  the  women 
contrasts  with  that  of  men  may  be  mentioned.  The  wau 
stresses,  in  his  naven,  emotions  which  are  freely  and  easily 
expressed  by  the  "mother "  whom  he  imitates,  but  the  women 
in  their  naven  exaggerate  special  aspects  of  the  father's  and 
elder  brother's  position  which  these  relatives  could  not  freely 
express  without  embarrassment.  The  father  must  repress  his 
disciplinary  activities  and  give  way  before  his  son,  but  the 


The  Expression  of  Ethos  in  Naven  215 

women  have  picked  upon  the  disciplinary  aspect  of  father- 
hood for  exaggeration  in  naven.  We  cannot  of  course  suppose 
that  the  repressed  emotions  of  a  father  are  a  dynamic  force 
in  the  shaping  of  the  women's  behaviour.  But  it  may  be  that 
the  women  get  some  extra  excitement  out  of  thus  exaggerating 
attitudes  which  are  to  some  extent  culturally  taboo  in  the 
relatives  whom  they  are  aping. 

At  the  beginning  of  this  chapter  I  have  already  indicated 
the  ethological  basis  of  the  women's  behaviour,  the  enjoy- 
ment which  they  get  out  of  their  swashbuckling  and  swagger, 
and  the  way  in  which  transvesticism  goes  hand  in  hand  with 
public  performance.  It  only  remains  therefore  to  consider 
the  symbolism  of  the  various  details : 

nyame  (mother).  The  behaviour  of  the  mother  we  have 
already  seen  to  be  a  simple  straightforward  expression  of  the 
latmul  mother's  pride  in  her  son.  There  is  in  her  case  no 
structural  basis  for  transvesticism,  and  her  feelings  about  her 
child  are  not  markedly  ambivalent;  therefore  there  is  in  her 
behaviour  none  of  the  elaboration  present  in  that  of  the  other 
relatives.  As  to  the  nakedness  of  the  mother  in  naven,  we 
may  compare  this  with  the  other  contexts  in  which  women 
are  ceremonially  naked,  namely,  during  the  burial  of  their 
nearest  male  relatives  (cf.  p.  154),  and  when  they  go  as 
suppliants  (cf.  p.  146).  It  may  seem  odd  that  the  mother 
should  adopt  exactly  the  same  gesture  both  to  express  distress 
and  to  rejoice  in  her  son's  achievement.  But  I  think  that  we 
may  regard  nakedness  in  all  these  three  contexts  as  an  extreme 
expression  of  abnegation  or  negative  self-feeling,  an  expres- 
sion which  may  be  accompanied  either  by  joy  or  sorrow. 

iau  (father's  sister).  The  naven  behaviour  of  the  iau  is 
based  upon  her  identification  with  her  brother,  the  father 
of  the  hero.  In  this  identification  there  are  clearly  two  possi- 
bilities which  might  be  stressed.  The  relationship  of  a  father 
to  his  son  is,  as  we  have  seen,  ambivalent;  the  father  being 
impelled  both  to  assert  his  authority  and  to  give  way  before 
his  son's  advancement.  Either  of  these  aspects  of  the  ambi- 
valent relationship  might  have  been  adopted  by  the  iau,  and 
she  has  actually  chosen  the  assertive  side  of  fatherhood ;  she 


2i6  The  Expression  of  Ethos  in  Naven 

puts  on  the  finest  of  homicidal  ornaments  and  beats  her 
"  son  ".  The  alternative  behaviour  would  have  been  to  drama- 
tise the  father's  retirement  at  his  son's  achievement.  We  may 
suppose  that  such  a  role  would  be  to  a  latmul  woman  rather 
uninteresting  and  inappropriate  to  her  transvesticism.  She 
would  be  adopting  male  costume  merely  to  play  a  role  more 
appropriate  to  female  ethos. 

There  is  so  far  as  I  know  no  detail  of  dramatisation 
referring  to  the  possibility  that  the  hero  may  marry  iau's 
daughter ;  and  the  absence  of  such  a  feature  may  be  noted  as 
additional  evidence  for  supposing  that  this  marriage  is  a 
recent  innovation  (see  footnote,  p.  90). 

tshaishi  (elder  brother's  wife).  The  behaviour  of  tshaishi 
is  based  upon  her  identification  with  her  husband,  the  hero's 
elder  brother;  and  to  a  very  great  extent  her  behaviour  and 
kinship  position  are  comparable  with  those  of  the  iau.  We 
have  already  seen  that  an  elder  brother  is  in  a  position 
analogous  to  that  of  father,  so  that  tshaishi  identifying  herself 
with  the  elder  brother  is  in  a  position  analogous  to  that  of  iau 
who  identifies  herself  with  the  father. 

The  only  special  feature  in  the  tshaishi's  position  may  be 
correlated  with  the  levirate  whereby  it  is  likely  that  she  will 
one  day  be  the  hero's  wife.  This  possibility  no  doubt  lends 
a  certain  piquancy  to  the  tshaishi  s  naven  and  is  exploited 
to  the  full  in  the  incident  in  which  the  hero  spears  a  fish  trap, 
the  symbol  of  his  tshaishi' s  womb.  The  same  aspect  of  the 
relationship  is  again  embroidered  upon  in  the  behaviour  of 
the  hero's  sister. 

nyanggai  (sister).  This  relative  plays  but  little  part  in  the 
naven,  but  she  is  identified  with  the  hero  on  the  occasion 
when  both  step  across  the  prostrate  women.  Then  the  sister 
expresses  with  exaggerated  dramatisation  what  we  (or  she) 
may  suppose  to  be  the  repressed  desires  of  the  hero.  She 
attacks  the  genitals  of  the  women  with  her  hands  and  especially 
those  of  tshaishi.  Upon  these  she  pounces  with  the  exclamation 
"A  vulva!"  but  the  transvestite  tshaishi  replies  "No!  A 
penis ! "  This  disagreement  between  them  is,  no  doubt,  due 
to  the  fact  that  the  sister  regards  herself  as  the  potential 


The  Expression  of  Ethos  in  Naven  217 

husband  of  tshaishi  while  the  tshaishi  regards  herself  as 
potentially  an  elder  brother  of  the  hero. 

mbora  (mother's  brother's  wife).  This  relative  identifies 
herself  with  the  wau.  Like  him  she  dances  with  the  captured 
head  (cf.  Plate  V  a)  and  receives  presentations  of  valuables 
from  the  laua.  Inasmuch  as  she  is  identified  with  a  man  who 
is  himself  transvestite,  it  is  not  surprising  to  find  that  the 
mbora  seems  to  be  in  some  doubt  about  her  sex,  so  that  she 
appears  in  naven  dressed  sometimes  as  a  man  and  sometimes 
as  a  woman.  Finally,  the  confusion  on  this  subject  would 
seem  to  be  resolved  when  she,  acting  as  a  man,  goes  through 
a  ritual  copulation  with  her  transvestite  husband. 

One  incident  remains  to  be  considered,  that  in  which  the 
iauSy  in  male  costume  and  carrying  a  feather  headdress,  come 
striding  across  the  prostrate  mboras  (also  in  male  costume). 
Of  this  I  can  give  only  a  tentative  explanation.  We  have  seen 
that  pride  in  a  man's  achievements  is  regarded  as  the  preroga- 
tive not  of  his  father's  clan,  to  which  he  himself  belongs,  but 
rather  of  his  mother's  clan.  It  would  seem  that  in  this  piece 
of  fooling  the  feather  headdress  is  a  homicidal  ornament.^ 
The  taus  (members  of  the  hero's  father's  clan)  come,  as  it 
were,  boasting  of  the  achievement.  They  stride  across  the 
mboras,  who  lie  on  the  ground,  as  the  mother  lies  before  her 
triumphant  son.  Both  their  posture  and  the  fact  that  mbora 
is  identified  with  wau  would  seem  to  indicate  that  the  mboras 
are  representatives  of  the  maternal  clan.  They  jump  to  their 
feet  and  snatch  the  feather  ornament,  the  symbol  of  achieve- 
ment, symbolically  claiming  it  as  the  triumph  of  the  maternal 
clan. 

A  piece  of  supporting  evidence  for  this  interpretation  may 
be  found  in  the  fact  that  homicidal  ornaments  are  usually 
presented  by  his  lanoa  nampa  (husband  people)  to  the  killer. 
This  would  be  consistent  with  the  interpretation  which  I  have 
put  forward. 

^  Unfortunately  my  informant  did  not  give  me  the  native  term  for 
the  particular  ornament  used  in  this  context.  He  only  stated  that  feathers 
were  stuck  into  pith,  the  usual  technique  in  making  certain  sorts  of 
homicidal  headdress.  It  is  probable,  however,  that  the  ornament  is  a 
badge  of  homicide. 


CHAPTER  XV 


The  Eidos  of  latmul  Culture 

FROM  the  foregoing  analysis  and  from  the  general  descrip- 
tions which  have  been  given  of  latmul  culture,  it  is 
evident  that  this  culture  has  some  general  pervasive 
character  other  than  that  which  it  derives  from  its  charac- 
teristic ethos.  It  is  a  culture  which  continually  surprises  us 
by  the  mass  of  structural  detail  which  it  has  built  up  around 
certain  contexts.  Most  conspicuously  we  have  the  fabric  of 
fancy  heraldry  and  totemism  built  up  around  the  personal 
names  and  ancestors :  but  a  similar  tendency  to  what  we  can 
only  describe  as  hypertrophy  may  be  recognised  in  the  initia- 
tory system,  with  its  plexus  of  cross-cutting  dual  divisions 
and  staggered  initiatory  grades ;  and  again  in  the  naven  cere- 
monial, where  we  have  seen  that  the  culture  has  proceeded 
upon  simple  structural  premises  to  such  lengths  that  the  wau 
behaves  as  in  some  sense  a  wife  of  the  laua. 

The  culture  has  apparently  some  internal  tendency  to 
complexity,  some  property  which  drives  it  on  to  the  fabrica- 
tion and  maintenance  of  more  and  more  elaborate  constructs, 
and  since  this  tendency  has  evidently  contributed  to  the 
shaping  of  naven,  it  is  relevant  to  examine  its  nature  in  detail. 
I  often  wondered  about  this  property  of  the  culture  when 
I  was  in  the  field,  but  never  saw  any  clue  to  an  understanding 
of  it.  It  was  only  after,  in  the  writing  of  this  book,  I  had 
worked  out  and  defined  cultural  structure  and  ethos,  that 
I  was  able  to  see  an  approach  to  the  present  problem.  I  did 
not  in  the  field  pay  any  special  attention  to  the  points  which 
I  know  now  might  have  thrown  light  on  this  problem,  but 
since  my  approach  has  been  but  little  attempted  by  others 
I  will  give  what  material  I  have  in  order  to  illustrate  the 
method. 

I  defined  cultural  structure  as  "a  collective  term  for  the 
coherent  'logical'  system  which  may  be  constructed  by  the 
scientist,  fitting  together  the  various  premises  of  the  culture  ", 


The  Eidos  of  latmul  Culture  219 

and  to  the  word  "logical"  I  added  a  footnote  to  the  effect 
that  we  must  expect  to  find  different  systems  of  logic,  dif- 
ferent modes  of  building  together  of  premises,  in  different 
cultures. 

The  first  point  in  this  definition  which  we  have  to  consider 
is  the  role  of  the  scientist.  I  said  that  cultural  structure  was 
a  system  built  up  by  the  scientist,  and  we  may  phrase  our 
present  problem  in  these  terms:  why  should  the  materials 
of  latmul  culture  drive  the  abstracting  scientist  to  greater 
complexities  of  exposition  than  are  demanded  by,  for  ex- 
ample, the  culture  of  the  Arapesh?^  Is  this  complexity  of 
exposition  the  reflection  of  a  real  complexity  in  the  culture, 
or  is  it  only  an  accidental  product  resulting  perhaps  from  a 
disparity  between  the  language  and  culture  of  the  ethno- 
grapher and  those  of  the  community  he  is  describing? 

Here  we  are  on  very  difficult  ground,  and  the  scientist 
who  has  set  out  the  exposition  is  perhaps  the  man  least 
fitted  to  judge  of  the  origin  of  its  complexity.  I  can  only 
say,  first,  that  the  latmul  culture  appeared  to  me  complex 
and  rich  before  I  even  attempted  a  structural  analysis,  and 
secondly,  that  the  result  of  analysis  has  been  to  cause  the 
culture  to  appear,  to  me  at  least,  not  more  complex,  but 
simpler.  I  can  only  trust  to  these  impressions  as  evidence 
that  the  complexity  is  not  entirely  a  creation  of  my  own 
methods  of  thinking.^ 

This  complexity,  then,  exists  in  some  form  in  the  culture 
itself,  and  is  built  up  of  premises.  The  word  premise  was 
defined  as  an  assumption  or  implication  recognisable  in  a 
number  of  details  of  cultural  behaviour,  and  I  mentioned 
that  though  these  assumptions  are  often  only  expressed  in 
symbolic  terms,  in  other  cases  they  are  articulately  stated  by 
the  natives. 

If  we  accept  {a)  that  the  complexity  of  the  cultural  struc- 
ture is  a  reflection  of  some  property  of  the  culture  itself,  and 
[h)  that  the  elements  of  this  complexity  are  to  some  extent 

^  Cf.  Margaret  Mead,  Sex  and  Temperament. 

^  To  some  extent  my  methods  of  thinking — structural,  ethological  and 
sociological — are  followed  by  the  natives  themselves.  This  matter  is  dis- 
cussed ar  the  end  of  this  chapter  (p.  250). 


220  The  Eidos  of  latmul  Culture 

present  as  ideas  or  assumptions  in  the  minds  of  the  natives, 
then  it  follows  that  any  pervasive  characteristic  of  the  cultural 
structure  can  be  referred  to  peculiarities  of  the  latmul  mind ; 
that,  in  fact,  we  are  here  dealing  with  the  cultural  expression 
of  cognitive  or  intellectual  aspects  of  latmul  personality. 

So  stated,  the  problem  appears  closely  analogous  to  that 
of  ethos.  In  each  case  we  have  to  account  for  a  generalised 
tendency  which  shows  itself  in  the  most  diverse  contexts  of 
the  culture,  and  it  would  seem  that  the  tendency  which  we 
are  now  investigating  bears  the  same  relation  to  cultural 
structure  as  that  which  ethos  bears  to  motivation.  We  have 
seen  that  ethos  constitutes  a  factor  in  the  determination  of 
the  needs  and  desires  of  the  individuals,  and  this  factor  varies 
from  culture  to  culture.  Ethos  is  the  system  of  emotional 
attitudes  which  governs  what  value  a  community  shall  set 
upon  the  various  satisfactions  or  dissatisfactions  which  the 
contexts  of  life  may  offer,  and  we  have  seen  that  ethos  may 
satisfactorily  be  regarded  as  "the  culturally  standardised 
system  of  organisation  of  the  instincts  and  emotions  of  indi- 
viduals ". 

In  an  analogous  way,  the  tendency  which  we  are  now 
examining  is  a  variable  factor  in  our  definition  of  cultural 
structure,  the  meaning  of  the  word  "logical",  the  factor 
which  determines  and  systematises  the  building  together  of 
structural  premises. 

From  the  close  analogy  between  the  two  problems  we  may 
guess  that  the  variable  factor  in  cultural  structure  is  some 
sort  of  standardisation  of  the  individuals  in  the  community. 
We  defined  social  structure  in  terms  of  ideas  and  assumptions 
and  "logic",  and  since  these  are,  in  some  sense,  a  product 
of  cognitive  process,  we  may  surmise  that  the  characteristics 
of  latmul  culture  which  we  are  now  studying  are  due  to 
a  standardisation  of  the  cogjiitive  aspects  of  the  personality  of 
the  individuals.  Such  a  standardisation  and  its  expression  in 
cultural  behaviour  I  shall  refer  to  as  the  eidos^  of  a  culture. 

^  The  sense  in  which  I  use  the  word  eidos  is,  of  course,  distinct  from 
that  in  which  it  was  used  in  Greek  philosophy,  but  since  the  philosophical 
term  has  not  been  adopted  in  this  form  into  the  English  language,  it  seems 
legitimate  to  use  the  word,  eidos,  in  the  present  sense. 

This  exposition  of  eidos  and  the  definition  of  social  structure  are, 


The  Eidos  of  latmul  Culture  221 

We  may  now  return  to  the  latmul  culture  and  examine  its 
eidos  in  some  detail.  Let  us  consider  first  the  phenomenon 
which  I  have  variously  described  as  hypertrophy,  structural 
complexity,  or  multiplicity  of  premises,  in  order  to  see 
whether  this  can  be  re-phrased  in  terms  of  the  standardisation 
of  cognitive  processes  of  individuals.  In  the  light  of  the 
definition  of  social  structure,  this  question  becomes  exces- 
sively simple.  Social  structure  is  built  up  of  "  ideas  or  assump- 
tions", and  the  culture  is  characterised  by  the  multiplicity 
of  elements  in  its  structure ;  that  is  to  say,  it  contains  a  very 
large  number  of  "ideas  or  assumptions",  products  of  cogni- 
tive process.  To  re-phrase  this  in  terms  of  the  standardisation 
of  individuals,  we  have  only  to  say  that  they  are  so  modified 
by  their  culture  that  their  output  of  cognitive  products  is 
increased.  In  fact,  that  they  are  stimulated  by  the  culture 
to  a  degree  of  intellectual  activity  unusual  among  primitive 
peoples. 

Here  I  am  not  referring  to  any  increase  in  intellectual 
endowment,  which  is  probably  controlled  to  a  great  extent  by 
hereditary  factors.  But  it  is  worth  mentioning  that  among 
the  latmul  definite  efforts  are  made  to  increase  the  memory 
endowment  of  individuals,  by  means  of  magical  techniques. 
Soon  after  birth,  a  male  child  is  made  to  inhale  smoke  from 
a  fire  which  has  been  bespelled,  in  order  that  the  boy  shall 
grow  up  to  be  erudite  in  the  totemic  names  of  his  clan ;  and 
later  in  life  a  man  may  be  treated  with  spells  which  are 
believed  to  act  on  his  heart  (the  seat  of  memory)  giving  him 
facility  in  the  memorising  of  name-cycles  and  spells.  In  a 
culture  in  which  such  techniques  have  been  evolved  it  is  at 
least  likely  that  there  is  a  factor  of  selection  which  favours 
those  of  high  cognitive  capacity.  In  the  present  instance, 
however,  I  am  referring,  not  to  an  increase  in  endowment, 
but  rather  to  the  way  in  which  the  cognitive  machinery  of 
the  individuals  is  stimulated  to  activity,  instead  of  being 
allowed  to  remain  idle. 

With  this  caveat — that  we  are  discussing  not  endowment 

I  think,  clumsy,  but  a  more  exact  formulation  of  the  matter  is  impossible 
until  an  adequate  enquiry  has  been  conducted  into  the  eidoses  and  ethoses 
■  of  various  cultures. 


222  The  Eidos  of  latmul  Culture 

but  stimulation — we  may  expect  to  find  that  cultures  differ 
enormously  in  the  extent  to  which  they  promote  intellectual 
activity :  though  I  know  of  no  field  work  in  which  any  atten- 
tion has  been  paid  to  this  phenomenon.  Any  mathematical 
evaluation  of  the  habitual  activity  of  the  intelligence  of  indi- 
viduals is,  of  course,  out  of  the  question.  But  there  are  many 
facts  which  indicate  that  a  high  degree  of  such  activity  is 
common  among  the  latmul  and  is  promoted  by  the  culture. 

The  present  state  of  our  knowledge  of  cognitive  aspects  of 
mental  process  is  so  unsatisfactory  that  it  is  not  possible  to 
describe  the  latmul  material  as  systematically  as  we  might 
desire.  We  may,  however,  very  roughly  classify  these  pro- 
cesses into  processes  of  recall  and  processes  of  thought,  both 
of  which  seem  to  have  contributed  to  the  maintenance  and 
development  of  complexity  in  the  latmul  culture. 

Let  us  first  consider  the  cultural  stimulation  of  memory.^ 
We  have  already  seen  that  vast  and  detailed  erudition  is  a 
quality  which  is  cultivated  among  the  latmul.  This  is  most 
dramatically  shown  in  the  debating  about  names  and  totems, 
and  I  have  stated  that  a  learned  man  carries  in  his  head 
between  ten  and  twenty  thousand  names.  This  figure  was 
arrived  at  by  very  rough  estimation  from  the  number  of  name 
songs  possessed  by  each  clan,  the  number  of  names  in  each 
song,  and  the  general  ability  of  such  men  to  quote,  in  con- 
siderable detail,  from  the  name-cycles  even  of  clans  other 
than  their  own.  The  figure  must  therefore  be  accepted  with 
caution,  but  it  is  certain  that  the  erudition  of  these  men  is 
enormous. 

Further  it  would  seem  that  rote  memory  plays  a  rather 
small  part  in  the  achievement  of  these  feats  of  memory.  The 
names  which  are  remembered  are  almost  all  of  them  com- 
pounds, each  containing  from  four  to  six  syllables,  and  they 
refer  to  details  of  esoteric  mythology,  so  that  each  name  has 
at  least  a  leaven  of  meaning.  The  names  are  arranged  in  pairs, 
and  the  names  in  any  one  pair  generally  resemble  each  other 
much  as  the  word  Tweedledum  resembles  the  word  Tweedle- 

^  I  have  been  much  influenced  in  my  own  thinking  about  these  pro- 
blems by  Professor  Bartlett's  book  Remembering,  1932,  which  I  read  after 
my  return  from  New  Guinea. 


The  Eidos  of  latmul  Culture  223 

dee — with  the  notable  difference  that  the  altered  syllable  or 
syllables  generally  have  some  meaning  and  are  connected 
together  by  some  simple  type  of  association,  e.g.  either  by 
contrast  or  by  synonymy.  A  progressive  alteration  of  meaning 
may  run  through  a  series  of  pairs. 

Thus  the  series  of  names  contain  tags  of  reference  which 
would  make  it  possible  for  them  to  be  memorised  either  by 
processes  of  imagery,  or  by  word  association.  I  collected 
a  great  quantity  of  these  names  and  noticed  again  and  again 
that  the  order  in  which  the  pairs  were  given  was  subject  to 
slight  but  continual  variation.  There  is  a  vaguely  defined 
standard  order  for  the  recitation  of  every  series  of  names. 
But  I  never  heard  any  criticism  of  the  order  in  which  names 
were  recited.  In  general,  an  informant  will  alter  the  order 
of  his  recitation  slightly  every  time  he  repeats  the  series. 
Occasionally  even,  the  pairing  of  the  names  is  altered,  but 
changes  of  this  type  are  definitely  regarded  as  mistakes. 

Bartlett^  has  pointed  out  that  one  of  the  most  characteristic 
qualities  of  rote  remembering  is  the  accuracy  with  which  the 
chronological  sequence  of  events  or  words  can  be  recalled. 
So  that  from  the  continual  alteration  of  the  order  in  which 
the  ixames  are  given  we  may  deduce  that  the  mental  process 
used  is  not  chiefly  that  of  rote  memory.  Additional  evidence 
for  this  conclusion  may  be  drawn  from  the  behaviour  of 
informants  when  they  are  endeavouring  to  recall  an  imper- 
fectly remembered  series  of  names.  I  do  not  remember  ever 
to  have  heard  an  informant  go  back,  like  a  European  child, 
to  the  beginning  and  repeat  the  series  of  names  already 
given,  in  the  hope  that  the  "impetus"  of  rote  repetition 
would  produce  a  few  more  names.  Usually  my  informants 
would  sit  and  think  and  from  time  to  time  produce  a  name 
(or  more  often  a  pair),  often  with  a  query  as  to  whether 
that  name  has  already  been  given — as  was  frequently  the 
case. 

Again,  when  a  latmul  native  is  asked  about  some  event 
in  the  past,  he  can  as  a  rule  give  an  immediately  relevant 
answer  to  the  question  and  does  not  require  to  describe  a 
whole  series  of  chronologically  related  events  in  order  to  lead 

^  Remembering,  pp.  203  and  264-266. 


224  The  Eidos  of  latmul  Culture 

up  to  the  event  in  question.  The  latmul  indulge  very  little 
in  the  sort  of  chronological  rigmarole  which,  as  Bartlett  has 
pointed  out,  is  characteristic  of  those  primitive  peoples  who 
have  specialised  in  rote  remembering. 

One  detail  of  the  culture  is  worth  mentioning  as  likely  to 
promote  the  higher  processes  rather  than  rote  memory.  This 
concerns  the  technique  of  debating.  In  a  typical  debate  a 
name  or  series  of  names  is  claimed  as  totemic  property  by 
two  conflicting  clans.  The  right  to  the  name  can  only  be 
demonstrated  by  knowledge  of  the  esoteric  mythology  to 
which  the  name  refers.  But  if  the  myth  is  exposed  and  be- 
comes publicly  known,  its  value  as  a  means  of  proving  the 
clan's  right  to  the  name  will  be  destroyed.  Therefore  there 
ensues  a  struggle  between  the  two  clans,  each  stating  that 
they  themselves  know  the  myth  and  each  trying  to  find  out 
how  much  their  opponents  really  know.  In  this  context,  the 
myth  is  handled  by  the  speakers  not  as  a  continuous  narrative, 
but  as  a  series  of  small  details.  A  speaker  will  hint  at  one 
detail  at  a  time — to  prove  his  own  knowledge  of  the  myth — 
or  he  will  challenge  the  opposition  to  produce  some  one  detail. 
In  this  way  there  is,  I  think,  induced  a  tendency  to  think  of  a 
story,  not  as  a  chronological  sequence  of  events,  but  as  a  set 
of  details  with  varying  degrees  of  secrecy  surrounding  each — 
an  analytic  attitude  which  is  almost  certainly  directly  opposed 
to  rote  remembering. 

But  though  we  may  with  fair  certainty  say  that  rote 
memory  is  not  the  principal  process  stimulated  in  latmul 
erudition,  it  is  not  possible  to  say  which  of  the  higher  pro- 
cesses is  chiefly  involved.  There  are,  however,  several  details 
of  the  culture  which  point  to  visual  and  kinaesthetic  imagery 
as  likely  to  be  of  great  importance.  In  debate,  objects  are 
continually  off"ered  for  exhibition.  For  example,  when  the 
totemic  ownership  of  the  Sepik  River  was  in  dispute,  a  shell 
necklace  was  hung  in  the  centre  of  the  ceremonial  house  to 
represent  the  river.  In  the  debating,  clan  A  claimed  that  the 
elephant  grass  which  forms  a  conspicuous  and  picturesque 
fringe  along  the  banks  of  the  river  was  indubitably  theirs; 
and  that  therefore  the  river  must  belong  to  them.  They 
accordingly  produced  a  beautiful  spear  decorated  with  leaves 


The  Eidos  of  latmul  Culture  225 

of  the  grass  and  pointed  to  it,  saying  "  Our  lambwiuishi ! !  "* 
Clan  B  on  the  other  hand  claimed  that  the  river  was  their 
snake,  Kindjin-kamboi,  and  their  protagonist,  MaH-kindjin 
went  off  to  get  brightly  coloured  leaves  to  ornament  a  repre- 
sentation of  his  snake  which  adorned  one  of  the  gongs  in  the 
ceremonial  house.  Again  in  a  debate  about  the  Sun,  a 
number  of  the  participants  dressed  themselves  up  to  repre- 
sent characters  in  a  myth  of  origin  of  the  Sun. 

In  the  technique  of  debating  the  speaker  uses  bundles  of 
leaves,  beating  on  a  table  with  them  to  mark  the  points  of 
his  speech.  These  leaves  are  continually  used  as  visible  or 
tangible  emblems  of  objects  and  names.  A  speaker  will  say, 
"This  leaf  is  So-and-so,  I  am  not  claiming  that  name",  and 
he  will  throw  the  leaf  across  to  the  opposition.  Or  he  may 
say,  "This  leaf  is  so-and-so's  opinion",  and  he  will  throw  it 
onto  the  ground  with  contempt ;  or  he  will  sweep  the  ground 
with  the  leaves,  brushing  away  his  opponents'  rubbishy  state- 
ments. Similarly  a  small  empty  leaf  packet  is  used  as  an 
emblem  of  some  secret,  of  which  the  speaker  is  challenging 
the  opposition  to  show  a  knowledge :  he  will  hold  it  up  asking 
them  scornfully  if  they  know  what  is  inside  it. 

The  proneness  to  visual  or  kinaesthetic  thought  is  shown 
too  in  the  continual  tendency  to  diagrammatise  social  organisa- 
tion. In  almost  every  ceremony,  the  participants  are  arranged 
in  groups  so  that  the  total  pattern  is  a  diagram  of  the  social 
system.  In  the  ceremonial  house  the  clans  and  moieties  are 
normally  allotted  seats  according  to  the  totemic  system  of 
groupings:  but  when  initiation  ceremonies  are  to  be  per- 
formed this  arrangement  is  discarded  and  in  its  place  comes 
another  based  upon  the  cross-cutting  initiatory  moieties  and 
grades.^ 

^  The  totemic  name  of  the  grass. 

*  For  other  material  of  this  kind,  see  Oceania,  1932,  passim;  the  state- 
ment on  p.  256  that  the  village  is  longitudinally  divided  between  the  two 
totemic  moieties  is,  however,  false.  It  was  collected  while  I  was  on  a 
short  visit  to  a  village  in  which  initiation  was  taking  place.  We  were  in 
the  initiatory  ceremonial  house,  and  it  would  appear  that  my  informant 
was  so  influenced  by  the  initiatory  patterning  inside  the  ceremonial  house 
that  he  spoke  as  if  that  patterning  extended  to  the  whole  village.  As  a 
matter  of  fact,  the  subdivision  of  the  village  does  on  the  whole  follow 
the  moiety  divisions,  but  is  transverse  and  not  longitudinal.  The  village 
of  Mindimbit  in  which  I  chiefly  worked  on  the  first  expedition  was 
irregularly  laid  out  owing  to  the  swampiness  of  the  ground. 


226  The  Eidos  of  latmul  Culture 

Lastly  we  may  cite  the  naven  ceremonies  as  a  further 
example  of  this  proneness  to  visual  and  kinaesthetic  thought. 
We  have  seen  how  the  abstract  geometrical  properties  of  the 
kinship  system  are  here  symbolised  in  costume  and  gesture ; 
and  we  may  note  this  in  passing  as  a  contribution  of  eidology 
to  our  understanding  of  the  ceremonies. 

But  the  connection  between  the  expression  of  eidos  in  the 
contexts  which  I  have  described,  and  the  culture  as  a  whole, 
is  still  not  perfectly  clear.  I  have  illustrated  the  eidos  chiefly 
from  the  totemic  debating,  and  have  shown  that  very  great 
activity  of  memory  is  demanded  and  promoted  in  certain 
individuals  by  the  sport  of  debating.  Further,  I  have  given 
facts  which  indicate  that  rote  memory  plays  only  a  small  part 
in  this  activity,  while  visual  and  kinaesthetic  imagery  appears 
to  be  important.  In  the  special  business  of  memorising  names 
I  showed  that  it  is  possible  to  suppose  that  word  association 
plays  a  part.  But  these  facts  might  well  be  isolated  in  their 
effects.  On  the  one  hand,  the  active  cultivation  of  memory 
might  be  confined  to  a  few  selected  specialists  and,  on  the 
other  hand,  it  might  occur  only  in  the  special  contexts  in 
which  names  are  important.  Until  these  two  possibilities 
have  been  examined,  we  cannot  step  from  the  facts  given 
to  the  statement  that  the  active  development  of  memory  has 
affected  the  culture  as  a  whole  and  the  naven  ceremonies  in 
particular. 

We  will  first  consider  how  far  this  activation  can  be  sup- 
posed to  have  affected  the  whole  community,  and  how  far 
it  is  confined  to  a  small  minority  of  specialists.  On  the  whole, 
the  remarkable  keenness  in  memorising  names  is  to  be  found 
in  the  majority  of  the  men.^  When  I  was  collecting  the  names 
I  got  my  material  as  far  as  possible  from  specialists,  but  it 
was  noticeable  that,  even  when  I  was  talking  about  other 
matters  with  informants  who  would  never  have  dared  to  pose 
as  erudite  in  public  debate,  they  would  continually  bring  the 
talk  round  to  matters  connected  with  the  totemic  system  and 
would  attempt  to  give  me  lists  of  names.  This  was  true,  for 

^  About  the  women  I  have  no  data,  but  from  the  contrast  in  ethos 
between  the  sexes  I  should  expect  to  find  considerable  differences  in 
their  eidos. 


The  Eidos  of  latmul  Culture  zz'j 

example,  of  the  informant  (Plate  XXII)  whom  I  have  described 
above  (p.  i6i)  as  conspicuously  enthusiastic  and  inaccurate. 
He  insisted  upon  discussing  esoterica  and  giving  lists  of 
names  belonging  to  his  clan,  full  of  blunders  and  contradic- 
tions. In  the  younger  men,  however,  this  passion  for  showing 
off  even  weakly  developed  erudition  is  almost  completely 
checked  by  the  feeling  that  erudition  is  only  appropriate  in 
senior  men.  I  had  three  very  intelligent  youths  who  con- 
sistently avoided  giving  me  names,  and  who  referred  me  to 
their  seniors  when  I  pressed  them.  But  I  was  told  by  other 
people  in  their  absence  that  two  of  these  youths  were  already 
well  on  the  way  towards  erudition,  and  would  be  great 
debaters  when  they  were  older.  Thus  the  reticence  of  the 
younger  men  on  the  subject  of  names  does  not  imply  that 
they  are  not,  like  their  seniors,  keen  on  this  form  of  mental 
virtuosity.  1 

But  a  more  complete  answer  to  the  question  of  how  the 
stimulation  of  a  small  number  of  specialists  can  react  on  the 
culture  as  a  whole  is  provided  by  the  fact  that  these  specialists 
constantly  set  themselves  up  as  unofficial  masters  of  cere- 
monies, criticising  and  instructing  the  men  who  are  carrying 
out  the  intricacies  of  the  culture.  Their  voice  is  heard  not 
only  in  the  debates  which  concern  totemic  names,  but  also 
in  those  on  every  subject  from  initiation  to  land  tenure.  Thus 
the  culture  is  to  a  great  extent  in  the  custody  of  men  trained 
in  erudition  and  dialectic  and  is  continually  set  forth  by 
them  for  the  instruction  of  the  majority.  From  this  we  may 
be  fairly  certain  that  the  individuals  most  affected  by  the 
stimulation  of  memory  actually  contribute  very  much  more 
than  their  fellows  to  the  elaboration  and  maintenance  of  the 
culture. 

Lastly,  there  is  the  question  whether  this  business  of 
totemic  names  constitutes  a  separate  context,  a  special  interest, 

^  The  reticence  of  the  younger  men  on  the  subject  of  names  may  be 
compared  with  their  comparative  quietness  in  the  ceremonial  house, 
a  matter  which  I  mentioned  in  the  discussion  of  the  men's  ethos. 
I  said  there  that  in  the  junior  ceremonial  house  the  boys  ape  the  violence 
of  their  seniors,  but  I  do  not  know  of  any  debating  in  the  junior  ceremonial 
house  in  which  the  erudition  of  the  older  men  is  copied.  There  are, 
however,  a  number  of  games  in  which  children  test  each  other's  know- 
ledge, e.g.  of  species  of  plants  in  the  bush,  etc. 


228  The  Eidos  of  latmul  Culture 

the  effects  of  which  might  be  isolated  from  the  culture  as  a 
whole;  or  whether  it  is  a  widely  ramifying  system  which 
might  be  expected  to  have  effects  in  all  parts  of  the  culture. 
Here  the  indications  are  very  strongly  in  favour  of  the  latter 
view.  The  naming  system  is  indeed  a  theoretical  image  of  the 
whole  culture  and  in  it  every  formulated  aspect  of  the  culture 
is  reflected.  Conversely,  we  may  say  that  the  system  has  its 
branches  in  every  aspect  of  the  culture  and  gives  its  support 
to  every  cultural  activity.  Every  spell,  every  song — even  the 
little  songs  which  boys  make  up  for  their  sweethearts — con- 
tains lists  of  names.  The  utterances  of  the  shamans  are 
couched  in  terms  of  names.  The  shamanic  spirits  which 
possess  the  shamans  are  themselves  important  nodal  points 
in  the  naming  system.  Marriages  are  often  arranged  in 
order  to  gain  names.  Reincarnation  and  succession  are  based 
upon  the  naming  system.  Land  tenure  is  based  on  clan 
membership  and  clan  membership  is  vouched  for  by  names. 
The  man  who  buys  names  acquires  at  the  same  time  member- 
ship in  the  clan  to  which  the  names  belonged,  and  a  right  to 
cultivate  the  land  of  that  clan.  Every  product  of  the  river 
and  gardens  has  its  place  in  the  system.  The  land,  the  river, 
the  sky  and  the  islands  of  floating  grass  are  all  named  in 
the  system.  Every  utensil  of  the  house,  the  house  itself  and 
the  ceremonial  house  are  included.  Every  man,  woman  and 
child  is  included,  and  the  various  aspects  of  social  personality 
are  differentiated  in  the  system  (cf.  infra). 

Indeed  the  only  province  of  the  culture  which  is  almost 
independent  of  the  naming  system  is  that  of  initiation;^  and 

^  Here  there  is  a  difference  between  the  Eastern  latmul  (Mindimbit, 
etc.)  and  the  Central  latmul  (Kankanamun,  etc.).  Among  the  latter, 
both  the  initiatory  moieties  and  the  totemic  moieties  are  called  "Sun" 
and  "Mother",  in  spite  of  the  fact  that  several  clans,  which  in  totemism 
belong  to  one  moiety  and  speak  on  that  side  in  debates  about  names, 
have  gone  across  to  the  other  moiety  for  initiatory  purposes.  But  in 
Mindimbit,  the  initiatory  dual  division  is  regarded  as  entirely  distinct 
from  the  totemic  and  the  initiatory  moieties  are  not  named  "Sun"  and 
"Mother".  They  are  called  Kishit  and  Miwot.  Moreover  in  Mindimbit 
the  secret  sHt  gongs  are  not  identified  with  shamanic  spirits.  In  Kankana- 
mun, both  gongs  and  spirits  are  called  zcagan;  but  in  Mindimbit  while 
the  gongs  are  still  called  wagan,  the  spirits  are  called  lemwail.  Thus  it 
would  seem  that  the  connection  between  totemism  and  initiation  is 
weaker  in  Mindimbit  than  in  Kankanamun. 


The  Eidos  of  latmul  Culture  229 

even  here  we  find  that  all  the  mysterious  sound-producing 
mechanisms  have  totemic  names,  the  secret  slit  gongs  are 
vaguely  equated  with  shamanic  spirits,  and  the  initiatory 
moieties  have  polysyllabic  names  on  the  usual  plan,  though 
no  place  has  been  allotted  to  these  in  the  distribution  of 
names  among  the  clans.  Lastly,  the  novices  receive  at  the 
end  of  initiation  special  names  from  their  mothers'  brothers. 

From  this  extraordinary  ramification  of  the  system  we  may, 
I  think,  be  fairly  certain  that  the  more  learned  men  approach 
every  cultural  activity  with  the  cognitive  habits  acquired  in 
debate ;  that  not  only  is  their  memory  activated  by  the  culture 
in  the  special  context  of  debating,  but  that  this  activation 
has  its  effects  on  the  culture  as  a  whole,  promoting  and  main- 
taining the  extreme  complexity  which  we  have  seen  to  be 
characteristic. 

Hitherto,  in  describing  latmul  eidos,  we  have  dealt  with 
its  quantitative  aspects  and  with  questions  as  to  what  types  of 
thought  mechanisms  are  most  cultivated.  It  now  remains  to 
consider  in  more  detail  the  patterning  of  the  standardised 
thought  and  the  sorts  of  logic  which  are  characteristic  of  it. 
Unfortunately  I  have  very  little  material  which  would  demon- 
strate the  methods  of  thought  of  individual  natives,  and  must 
therefore  depend  almost  entirely  upon  the  details  of  the 
culture,  deducing  therefrom  the  patterns  of  thought  of  the 
individuals.  Ideally  it  should  be  possible  to  trace  the  same 
processes  in  the  utterances  of  informants  and  in  individual 
behaviour  in  experimental  conditions  as  well  as  in  the  norms 
of  the  culture. 

In  the  first  place,  we  must  note  that  a  great  deal  of  latmul 
thought  is  "intellectual" — and  here  I  am  using  the  term  in 
that  slang  sense  in  which  it  is  used  to  imply  that  people  are 
"intellectual"  rather  than  intelligent. 

The  problems  which  most  exercise  the  latmul  mind  appear 
to  us  fundamentally  unreal.  There  is,  for  example,  a  standing 
argument  between  the  Sun  moiety  and  the  Mother  moiety 
as  to  the  nature  of  Night.  While  the  Sun  people  claim  Day 
as  their  totemic  property,  the  Mother  people  claim  Night 
and  have  developed  an  elaborate  esoteric  rigmarole  about 
mountains  meeting  in  the  sky,  and  ducks,  and  the  Milky  Way 


230  The  Eidos  of  latmul  Culture 

to  explain  its  existence.  The  Sun  people  are  contemptuous 
of  this  and  Night  has  become  a  bone  of  contention.  The 
Mother  people  maintain  that  Night  is  a  positive  phenomenon 
due  to  the  overlapping  mountains,  etc.,  while  the  Sun  people 
maintain  that  Night  is  a  mere  nullity,  a  negation  of  Day, 
due  to  the  absence  of  their  totem,  the  Sun. 

Again,  within  the  Sun  moiety  itself  there  is  matter  for 
dispute  about  the  Sun.  One  of  his  totemic  names  is  Twat- 
mali,i  but  there  are  two  clans  in  this  moiety  who  lay  claim 
to  this  name.  Each  clan  has  its  own  string  of  names  linked 
with  this.  Thus  clan  A  claims  that  the  series  runs :  Twat-mali, 
Awai-mali ;  Ka-ruat-mali,  Kisa-ruat-mali ;  etc.  (nine  pairs  of 
names),  while  clan  B  claims  that  the  series  runs :  Twat-mali, 
Awai-mali ;  Ndo-mbwangga-ndo,  Kambwak-mbwangga-ndo ; 
etc.  (eight  pairs  of  names).  At  some  time  in  the  past  a  settle- 
ment of  this  argument  has  apparently  been  reached  in  a  com- 
promise: that  there  are  two  Twat-malis,  one  of  whom  is 
the  sun  who  shines  nowadays,  while  the  other  is  the  old  sun, 
who  lies  as  a  decaying  rock  somewhere  in  the  plains  north 
of  the  Sepik  River.  But  the  settlement  is  only  partial,  because 
it  has  never  been  agreed  which  of  the  two  Twat-malis  is 
which,  and  nowadays  each  clan  taunts  the  other  by  saying 
that  their  own  Twat-mali  is  in  the  sky,  while  that  of  the 
opposition  lies  rotting  in  the  plains. 

Another  subject  which  is  matter  for  this  characteristic 
intellectual  enquiry  is  the  nature  of  ripples  and  waves  on  the 
surface  of  water.  It  is  said  secretly  that  men,  pigs,  trees, 
grass — all  the  objects  in  the  world — are  only  patterns  of 
waves.  Indeed  there  seems  to  be  some  agreement  about  this, 
although  it  perhaps  conflicts  with  the  theory  of  reincarnation, 
according  to  which  the  ghost  of  the  dead  is  blown  as  mist  by 
the  East  Wind  up  the  river  and  into  the  womb  of  the  de- 
ceased's son's  wife.  Be  that  as  it  may — there  is  still  the 
question  of  how  ripples  and  waves  are  caused.  The  clan 
which  claims  the  East  Wind  as  a  totem  is  clear  enough  about 
this :  the  Wind  with  her  mosquito  fan  causes  the  waves.   But 

^  Other  clans  of  the  Sun  moiety  have  other  names  for  the  Sun — 
langgun-mah,  Kala-ndimi,  etc.  The  dispute  here  discussed  is  concerned 
merely  with  Twat-mali. 


The  Eidos  of  latmul  Culture  23 1 

other  clans  have  personified  the  waves  and  say  that  they  are 
a  person  (Kontum-mali)  independent  of  the  wind.  Other 
clans,  again,  have  other  theories.  On  one  occasion  I  took 
some  latmul  natives  down  to  the  coast  and  found  one  of 
them  sitting  by  himself  gazing  with  rapt  attention  at  the  sea. 
It  was  a  windless  day,  but  a  slow  swell  was  breaking  on  the 
beach.  Among  the  totemic  ancestors  of  his  clan  he  counted 
a  personified  slit  gong  who  had  floated  down  the  river  to 
the  sea  and  who  was  believed  to  cause  the  waves.  He  was 
gazing  at  the  waves  which  were  heaving  and  breaking  when  no 
wind  was  blowing,  demonstrating  the  truth  of  his  clan  myth. 

On  another  occasion  I  invited  one  of  my  informants  to 
witness  the  development  of  photographic  plates.  I  first  de- 
sensitised the  plates  and  then  developed  them  in  an  open 
dish  in  moderate  light,  so  that  my  informant  was  able  to  see 
the  gradual  appearance  of  the  images.  He  was  much  inter- 
ested, and  some  days  later  made  me  promise  never  to  show 
this  process  to  members  of  other  clans.  Kontum-mali  was 
one  of  his  ancestors,  and  he  saw  in  the  process  of  photo- 
graphic development  the  actual  embodiment  of  ripples  into 
images,  and  regarded  this  as  a  demonstration  of  the  truth  of 
the  clan's  secret. 

This  intellectual  attitude  towards  the  great  natural  phe- 
nomena crops  up  continually  in  conversations  between  the 
anthropologist  and  his  informants,  the  latter  striving  to  pump 
the  anthropologist  about  the  nature  of  the  universe.  I  learned 
very  soon  that  the  correct  attitude  to  adopt  in  such  conversa- 
tions was  one  of  extreme  discretion.  I  parted  with  informa- 
tion only  after  insisting  upon  secrecy  and  then  shared  the 
secret  as  a  bond  between  my  informant  and  myself.  With 
such  preliminaries,  what  I  said  was  treated  seriously,  but 
without  them  it  was  generally  assumed  to  be  exoteric  lies. 
One  man  came  quietly,  boasting  to  me  that  he  knew  the 
European  secrets  about  day  and  night ;  that  a  white  man  had 
told  him  that  by  day  the  Sun  travelled  over  the  earth  and 
then  during  the  night  he  returned  to  the  East,  travelling  back 
over  the  sky- world,  so  that  the  people  in  the  sky  have  day 
while  we  on  earth  are  having  night  and  vice  versa — a  pretty 
re-phrasing  of  the  Antipodes  in  latmul  terms. 


232  The  Eidos  of  latmul  Culture 

A  lunar  eclipse  occurred  while  I  was  in  Kankanamun  and 
I  discussed  this  with  my  informants.  I  was  rather  surprised 
to  find  them  but  little  interested  in  the  phenomenon.  They 
ascribed  it  to  magic  performed  by  the  Tshuosh  people.  One 
informant,  lowimet,  seemed  rather  doubtful  about  this  theory, 
and  asked  me  what  we  Europeans  believed.  I  hedged  for  a 
while,  and  then  explained  the  secret.  The  moon  during  the 
eclipse  had  turned  red,  and  my  informant  had  told  me  some 
days  previously  that  the  sun  was  a  cannibal.  I  ascribed  the 
redness  of  the  moon  to  blood  in  the  sun's  excrement,  and 
said  that  this  material  was  between  the  moon  and  the  earth 
obscuring  the  moon's  brightness.  This  secret  was  quite  a 
useful  bond  of  understanding  between  my  informant  and  me, 
and  several  times  later  he  referred  to  it  as  "that  matter  we 
both  know  about".  It  might  be  interesting  to  collect  in  the 
future  lowimet's  re-phrasing  of  what  I  told  him. 

More  specifically,  latmul  thought  is  characterised  not  only 
by  its  intellectuality,  but  also  by  a  tendency  to  insist  that 
what  is  symbolically,  sociologically,  or  emotionally  true,  is 
also  cognitively  true.  Apparently  the  sort  of  paradox  which 
can  be  constructed  in  this  way  is  very  attractive  to  the  latmul 
mind,  and  the  same  mental  twist  is,  of  course,  recognisable 
in  dialecticians  and  theologians  in  all  parts  of  the  world. 
Among  the  latmul  the  dialecticians  and  theologians  are  not 
a  class  apart  but  are,  as  we  have  seen,  the  chief  contributors 
to  the  culture.  Thus  it  comes  about  that  many  of  the  com- 
plications of  the  culture  can  be  described  as  tours-de-force 
played  upon  this  type  of  paradox,  devices  which  stress  the 
contradiction  between  emotional  and  cognitive  reality  or  be- 
tween diff"erent  aspects  of  emotional  truth. 

As  examples  of  this,  we  have  the  oft-repeated  idea  of  the 
discrimination  between  and  yet  the  identity  of  different 
aspects  of  social  personality.  We  have  already  noted  this 
cultural  phenomenon  in  the  naming  of  human  beings;  we 
have  seen  how  a  man's  various  names  constitute  a  diagram 
of  his  various  allegiances  to  his  paternal  clan,  his  maternal 
clan,  and  his  classificatory  zt'«ws,  and  how  another  name  is 
connected  with  his  status  as  an  initiate.  The  various  souls  of 
a  man  with  their  different  fates  after  death — his  patrilineal 


The  Eidos  of  lattnul  Culture  233 

aspect  which  is  reincarnated  in  his  son's  son,  his  matrilineal 
aspect  which  dwells  on  in  the  land  of  the  dead  under  the 
name  given  him  by  his  wau,  and  that  aspect  of  his  personality 
which  is  represented  by  his  secret  name,  which  becomes  a 
potsherd  spirit,  a  guardian  of  his  clan — correspond  in  the 
same  way  with  the  various  facets  of  his  personality. 

The  discrimination  of  facets  of  personality  is  carried  even 
farther  in  the  case  of  wagan,  the  most  important  type  of 
spiritual  being.  Kava-mbuangga,  the  wagan  who  set  his  foot 
upon  the  mud  and  thereby  created  dry  land,  is  a  representa- 
tive both  of  the  living  and  of  the  dead,  and  there  are  myths 
of  his  visit  to  the  land  of  the  dead.  A  figure  of  him  which 
I  was  shown  in  Mindimbit  was  painted  in  two  colours, 
the  right-hand  side  of  the  figure  in  ochre  to  represent 
living  flesh,  and  the  left-hand  side  in  black  to  represent 
stone,  the  kava  (paralysis,  or  "  pins-and-needles ")  which 
creeps  over  the  body  in  death.  Thus  in  the  figure  the  two 
sides  of  Kava-mbuangga's  personality  are  diagrammatically 
shown. 

In  their  naming,  too,  the  wagan  are  differentiated,  not  only 
qua  members  of  their  maternal  and  paternal  clans  and  as 
initiates,  but  also  they  have  separate  strings  of  names  according 
as  they  function  as  {a)  shamanic  spirits ;  {b)  avenging  spirits, 
invoked  in  sorcery;  {c)  secret  slit  gongs  in  initiatory  cere- 
monies ;  {d)  mhwatnggowi  (ceremonial  dolls  which  are  vaguely 
believed  to  be  associated  with  fertility) ;  {e)  mwai  (long-nosed 
masks  in  which  the  young  men  dance.  These  dances  are 
junior  analogues  of  the  wagan  slit-gong  ceremonies  performed 
by  the  senior  men). 

These  different  names  of  the  wagan  are  many  of  them 
esoteric,  and  especially  to  be  concealed  is  the  fact  of  funda- 
mental identity,  which  exists  in  spite  of  the  differentiation 
between  the  various  personifications.  Only  the  more  erudite 
men  know  that  mwai  and  mhwatnggowi  are  really  wagan ;  and 
when  they  told  me  of  this  they  spoke  of  it  as  of  a  mystery 
which  was  beyond  their  comprehension,  a  paradox  which  had 
to  be  accepted  and  at  which  they  marvelled  with  a  certain 
serious  humility  and  acceptance  of  its  incomprehensibility. 
At  the  same  time,  they  insisted  upon  the  identity  with  an 


234  ^^^  Eidos  of  latmul  Culture 

over-emphasis  in  favour  of  that  behef  which,  of  the  two, 
was  the  more  difficuh  to  assimilate. 

Another  example  of  insistence  upon  a  sociological  truth 
because  it  looks  like  cognitive  nonsense  may  be  seen  in  the 
constantly  repeated  statements  that  the  Borassus  palm,  Tep- 
meaman,^  is  a  fish  as  well  as  a  palm.  Tepmeaman  is  the  kaishi 
(counterpart  or  partner)  of  nau,  the  sago  palm.  But,  in 
latmul  economics,  the  counterpart  of  sago  is  fish,  since  the 
river  people  catch  fish  which  they  exchange  for  sago  manu- 
factured in  the  bush  villages.  Thus  we  can  detect  some  justifi- 
cation for  the  statement  that  the  Borassus  palm  is  a  fish. 

The  tendency  to  emphasise  the  more  cognitively  obscure 
of  tAvo  truths  is  recognisable  again  in  the  kinship  system. 
In  my  preliminary  account  of  the  culture  I  stated  that  while 
the  "morphology  of  the  social  system "^  is  patrilineal,  the 
"sentiment"  of  the  people  is  preponderantly  matrilineal. 
I  am  inclined  now  to  see  in  this  almost  sentimental  stressing 
of  the  importance  of  the  tie  with  the  mother  another  instance 
of  over-compensation  in  favour  of  the  less  evident  truth. 
My  informants,  when  we  were  discussing  the  mother  and 
mother's  brother,  had  certainly  a  trace  of  that  same  puzzled 
insistence  which  was  present  when  they  told  me  of  the  dif- 
ferent personalities  of  wagan.  It  would  seem  then  that  the 
patrilineal  relationship  is  evident  enough  from  clan  organisa- 
tion to  need  no  stressing.  But  its  clear  existence  casts  a  slight 
obscurity  upon  the  matrilineal  relationship  which,  though  no 
less  socially  and  emotionally  real,  is  thereby  rendered  less 
cognitively  evident.  Therefore  the  matrilineal  relationship 
is  the  more  emphasised. 

*  The  names  of  this  palm  are  Tepmeaman,  Kambuguli,  etc.  (cf.  p.  44, 
where  the  complete  list  is  given).  In  a  previous  publication  {Oceania, 
1932)  I  have  stated  that  Kambuguli  is  a  different  species  of  palm 
from  Tepmeaman.  On  my  second  expedition  I  found  that  borassus  is 
a  dioecious  palm  and  that  Kambuguli  is  the  male  plant  as  distinct  from 
Tepmeaman,  the  female.  Actually  the  natives  regard  Tepmeaman  as  the 
male,  the  word  tshik  being  used  homonymously  for  "fruit"  and  "penis". 

"  Oceania,  1932,  p.  289.  By  "the  morphology  of  the  system"  I  meant 
those  cultural  premises  which  are  articulately  formulated  by  the  natives, 
e.g.  groupings  such  as  clans,  moieties,  initiatory  grades,  etc.,  which  are 
named  and  discriminated.  A  man's  ties  with  his  father  are  articulately 
stated,  while  those  with  his  mother  are  chiefly  expressed  in  symbolic 
behaviour. 


The  Eidos  of  latmul  Culture  235 

Another  example  of  the  same  sort  of  thinking  occurs  in 
the  identification  of  the  laua  with  the  ancestors  of  the  maternal 
clan  (cf.  p.  43).  Here  the  link  upon  which  the  equation  is 
based  would  seem  to  be  a  purely  emotional  one.  In  the  two 
relationships,  that  between  a  clan  and  its  lauas,  and  that 
between  a  clan  and  its  ancestors,  the  only  common  factor  is 
the  pride  which  is  associated  with  the  achievements  of  both 
lauas  and  ancestors. 

In  the  naven  ceremonies  we  can  see  the  same  sort  of  twisted 
thinking  acted  out  in  ceremonial.  In  a  social  and  economic 
and  emotional  sense,  the  wau  is  a  mother  and  wife  of  his  laua. 
But  in  a  strictly  cognitive  sense  he  is  not  anything  of  the  sort. 
In  the  carrying  out  of  the  ceremonial,  it  is  the  emotional 
truth  which  is  stressed,  which,  of  the  two,  is  the  more  diffi- 
cult of  cognitive  assimilation. 

Besides  this  sense  of  the  contrast  between  emotional  and 
cognitive  truth,  latmul  thought  has  certain  other  patterns 
or  motifs  in  which  it  is  trained  by  the  culture  and  which 
pervade  the  various  institutions.  At  this  stage  in  the  book 
it  will  hardly  surprise  the  reader  to  be  told  that  these  patterns 
are  to  a  great  extent  mutually  contradictory,  so  that  the 
culture  as  a  whole  appears  as  a  complex  fabric  in  which  the 
various  conflicting  eidological  motifs  are  twisted  and  woven 
together.  Unravelling  and  isolating  these  motifs,  we  can 
recognise  and  describe  some  of  the  more  important : 

{a)  A  sense  of  pluralism:  of  the  multiplicity  and  differ- 
entiation of  the  objects,  people,  and  spiritual  beings  in  the 
world. 

(6)  A  sense  of  monism ;  that  everything  is  fundamentally 
one  or  at  least  is  derived  from  one  origin. 

{c)  A  sense  of  direct  dualism :  that  everything  has  a  sibling. 

{d)  A  sense  of  diagonal  dualism — that  everything  has  a 
symmetrical  counterpart. 

{e)  Patterns  of  thought  which  govern  the  seriation  of  indi- 
viduals and  groups;  these  patterns  being  apparently  based 
upon  [c]  and  {d). 

The  senses  of  multiplicity  and  of  monism  need  only  very 
brief  treatment,  since  neither  of  them,  so  far  as  I  know,  has 
had  any  repercussions  on  the  naven  ceremonies.   On  the  one 


236  The  Eidos  of  latmul  Culture 

hand,  we  have  the  type  of  material  which  I  have  already 
mentioned,  the  distribution  of  thousands  of  personal  names 
to  the  things  and  people  of  the  world  and  to  the  different 
facets  of  their  personalities ;  and  on  the  other  hand,  we  have 
the  insistence  that  everything  is  really  one.  On  the  whole,  it 
would  seem  that  the  belief  in  multiplicity  is  lay  and  exoteric, 
while  the  details  of  the  basis  of  monism  are  an  esoteric 
mystery  for  the  erudite. 

This  does  not  mean  that  the  general  concept  of  monism 
is  a  secret.  When  two  clans  are  striving  over  the  ancestral 
ownership  of  a  name,  a  man  may  declaim  in  open  debate: 
"  One  father,  one  mother,  one  root — that  is  the  sort  of  people 
that  we  (human  beings)  are",  insisting,  by  this  conventional 
reference  to  one  of  the  pairs  of  first  parents  of  the  human  race, 
upon  the  indivisibility  of  the  disputed  name. 

But  the  names  of  these  first  parents  are  secret,  as  also  are 
all  the  details  of  muddled  mythology  whereby  monism  unites 
the  whole  world  in  historical  mysticism. 

The  same  considerations  apply  to  wagan.  There  are  many 
wagan  and  many  of  them  have  multiple  personalities,  and 
yet  all  wagan  are  one  and  were  descended  from  Kava- 
mbwangga  who  multiplied  himself  by  means  of  boluses  of 
chewed  betel  upon  which  he  trod  and  which  thereby  became 
the  other  zcagan;  but  which  are  all  really  Kava-mbwangga. 
Another  instance  of  this  mystical  monism  is  to  be  seen  in  the 
theory  that  all  things  in  the  world  are  only  the  patterns  of 
waves. 

But  over  against  the  sense  of  the  unity  of  everything,  we 
may  set  the  confusion  which  has  perhaps  resulted  from  the 
rivalry  between  clans  and  the  emphasis  upon  pride.  Every- 
body believes  and  insists  on  fundamental  monism,  but  each 
group  of  clans  has  its  own  mythological  theory  according  to 
which  its  own  ancestors  are  given  the  key  positions  in  the 
unitary  origin  of  the  world.  Thus  it  has  come  about  that  there 
is  not  one  monistic  theory  but  a  whole  series  of  conflicting 
theories,  each  of  which  emphasises  the  fundamental  unity 
of  the  world. 

There  are  at  least  two  (and  probably  more)  pairs  of  first 
parents  included  in  the  polysyllabic  totemic  scheme;  and 


The  Eidos  of  latmul  Culture  237 

besides  these  there  is  a  third  pair  who  have  short  names  and 
are  attached  to  no  clan^ — and  so  the  cuhure  swings  between 
monism  and  pluraHsm. 

An  interesting  expression  of  the  fundamental  pluraHsm  of 
the  world  may  be  seen  in  the  native  theory  of  separate  "  roads  ". 
This  is  a  matter  upon  which  the  natives  themselves  are  vague 
and  opinion  varies  from  one  informant  to  another.  I  was  told 
that  human  beings,  wagan,  kurgwa  (witches),  and  windjitnbu 
(wood  spirits),  have  all  of  them  separate  "  roads  ".  But  some 
informants  were  inclined  to  think  that  there  were  only  two 
"roads",  that  of  human  beings  and  that  of  spirits;  others 
again  discriminated  three  roads,  that  of  wagan  and  kurgwa^ 
that  of  wtndjimbu,  and  that  of  human  beings.  I  was  told  in 
pidgin  English  that  wagan  were  "behind  true",  i.e.  that 
though  invisible  they  were  yet  present  in  some  mysterious 
way  which  we  might  express  by  a  reference  to  the  "fourth 
dimension"  or  to  another  "plane  of  existence".  The  word, 
iamha  (road),  as  applied  to  these  spiritual  beings  seemed, 
although  the  roads  were  described  as  means  of  instantaneous 
transport,  to  be  also  an  equivalent  of  "planes  of  existence". 

While  the  monism  and  pluralism  in  latmul  thought  are 
comparatively  easy  to  understand,  the  two  forms  of  dualism 
are  somewhat  more  obscure.  The  matter  is,  however,  of  some 
importance,  and  I  believe  that  the  discrimination  of  direct 
and  diagonal  dualism  may  prove  useful  in  the  description  of 
cultures  other  than  the  latmul.^ 

^  It  is  interesting  that  my  informant,  Kainggenwan  of  Mindimbit,  who 
gave  me  the  names  of  these  two  non-totemic  first  parents,  Mogavia  (the 
man)  and  Leren  (the  woman),  regarded  these  names  as  secrets  of  great 
importance ;  although  as  they  have  no  clan  status  and  no  magical  efficacy 
we  might  expect  the  names  to  have  no  value. 

^  This  exposition  of  direct  and  diagonal  dualism  was  written  before 
I  realised  that  these  concepts  are  only  eidological  analogues  of  comple- 
mentary and  symmetrical  ethos.  The  phrasings  which  are  here  used  are 
in  some  respects  clumsy,  and  some  simplification  might  have  been  intro- 
duced by  substituting  the  term  complementary  for  "direct"  and  sym- 
metrical for  "diagonal".  It  seemed  worth  while,  however,  to  leave  the 
original  statement  of  the  matter  almost  unchanged  so  that  the  reader 
might  see  how  schismogenic  phenomena  were  reflected  in  my  account 
of  latmul  eidos  before  I  had  developed  the  concept  of  schismogenesis. 
The  equivalence  between  the  two  types  of  dualism  and  the  two  types  of 
schismogenesis  is  worked  out  more  fully  at  the  end  of  the  book,  since  it 
was  only  in  the  writing  of  the  Epilogue  that  this  identity  became  clear. 


238  The  Eidos  of  latmul  Culture 

By  dualism  I  mean  a  tendency  to  see  things,  persons,  or 
social  groups  as  related  together  in  pairs;  and  according  to 
the  type  of  relationship  which  is  seen  between  the  elements 
in  each  pair,  I  shall  speak  of  the  dualism  as  being  either  direct 
or  diagonal.  In  the  case  of  direct  dualism  the  relationship 
is  seen  as  analogous  to  that  which  obtains  between  a  pair  of 
siblings  of  the  same  sex ;  while  in  diagonal  dualism  the  rela- 
tionship is  seen  as  analogous  to  that  between  a  pair  of  men 
who  have  married  each  other's  sisters. 

These  considerations  were  suggested  to  me  by  a  paper  read 
by  Dr  Leonard  Adam,^  who  mentioned  a  Nepalese  institution 
which  he  described  as  "artificial  brotherhood".  The  exis- 
tence of  two  forms  of  dualist  thinking  became  evident  from 
considering  the  contrast  between  this  relationship  and  the 
corresponding  "artificial"  relationships  institutionalised  in 
latmul  culture. 

In  almost  all  cultures  there  are  to  be  found  these  institu- 
tionalised partnerships,  and  in  the  simpler  societies  it  is  usual 
to  find  that  these  partnerships,  contracted  either  outside  or 
across  the  bonds  of  kinship,  are  yet  patterned  upon  the 
familiar  kinship  ties.  Thus  in  Nepal  the  behaviour  of  the 
partners  is  apparently  patterned  upon  that  normal  between 
brothers  in  Nepalese  culture.  Among  the  latmul,  on  the 
other  hand,  there  is  no  "artificial  brotherhood",  but  in  its 
place  we  find  a  series  of  types  of  "artificial"  relationships 
patterned  upon  affinal  ties  and  stressing  the  symmetrical 
nature  of  these  ties. 

The  contrast  is  very  striking  and  led  naturally  to  the 
problem  of  what  cultural  factors  could  be  expressed  by 
institutionalising  the  one  type  of  "artificial"  relationship 
rather  than  the  other.  The  theory  which  I  am  now  putting 
forward  is  an  attempt  to  answer  this  question  from  an  eido- 
logical  point  of  view. 

The  theory  is  simply  that  there  are  two  contrasting  types 
of  dualist  thinking.  The  first  type,  which  is  easy  for  Euro- 

^  Man,  vol.  xxv,  1935,  p.  12  and  Zeitschr.  fur  Vergl.  Rechtswissenschaft, 
Bd.  XLix.  For  instances  of  "artificial  relationships"  probably  based  upon 
diagonal  dualisnm  in  African  communities,  cf.  Hocart,  "  Blood  Brother- 
hood", Man,  1935,  p.  127. 


The  Eidos  of  I attnul  Culture  239 

peans  to  follow,  would  classify  together  the  adjacent  corners 
of  a  rhombus  (to  employ  a  geometrical  analogy).  This  type  of 
thought,  which  I  shall  refer  to  as,  direct  dualism,  is  that  which 
leads  to  the  formation  of  artificial  relationships  based  upon 
brotherhood  and  to  concepts  such  as  that  of  the  latmul  that 
everything  in  the  world  can  be  grouped  in  pairs,  such  that 
in  each  pair  one  component  is  an  elder  sibling,  while  the 
other  is  a  younger  sibling  of  the  same  sex. 

The  second  type  of  thought  is  much  more  difficult  for 
Europeans  because  it  is  less  developed  in  our  eidos. ^  It  would 
classify  together  the  diagonally  opposite  corners  of  a  rhombus, 
stressing  their  fundamental  equality  and  oppositeness.  This 
diagonal  way  of  thinking  leads  in  latmul  culture  to  the  forma- 
tion of  artificial  affinal  relationships  and  to  the  idea  that 
everything  in  the  world  has  its  equal  and  opposite  counter- 
part. It  is  extended  farther  to  the  great  dualist  constructs 
underlying  the  two  moieties  of  Sun  and  Mother  with  their 
opposed  totems  Sky  and  Earth,  Day  and  Night,  and  to  the 
cross-cutting  dual  divisions  of  the  initiatory  system  in  which 
one  half  of  one  moiety  initiates  the  diagonally  opposite  half 
of  the  opposite  moiety  (see  Fig.  4,  p.  245). 

It  is  possible  that  these  two  patterns  of  thought  are  both 
of  them  represented  in  the  eidos  of  all  cultures,  but  I  believe 
that  there  is  a  great  deal  of  variation  between  cultures  in  the 
degree  to  which  one  or  the  other  type  of  dualism  is  stressed. 
Among  the  latmul,  both  types  are  strongly  developed.  It  is 
hard  to  estimate  their  relative  importance  because  the  diagonal 
way  of  thinking  appears  more  strange  and  striking  to  the 
European  observer.  My  impression  is  that  greater  emphasis 
is  given  to  the  diagonal  type,  and  further  that  this  emphasis 
is  stronger  among  the  Eastern  latmul  (Mindimbit,  Tam- 
bunum,  etc.)  than  among  the  Central  latmul  (Palimbai, 
Malingai,  etc.).  In  Mindimbit  I  collected  two  sorts  of  lists 
of  names  of  shamanic  spirits  arranged  in  pairs.    In  one  list 

^  It  may  reasonably  be  asked  why,  since  both  kinds  of  schismogenesis 
are  present  in  European  society,  it  should  be  more  difficult  for  us  to  think 
in  terms  of  diagonal  (i.e.  symmetrical)  dualism.  The  answer  is  that  while 
we  habitually  think  of  complementary  relationships  in  dual  terms,  we 
do  not  think  of  rivalry  and  competition  as  necessarily  dual.  Our  society, 
though  it  is  based  upon  competition  and  rivalry,  lacks  bilateral  symmetry. 


240  The  Eidos  of  latmul  Culture 

the  pairs  were  composed  of  elder  brother  and  younger  brother 
names,  while  in  the  other  list  they  were  pairs  of  kaishi  (oppo- 
site counterparts).  In  Palimbai  I  tried  to  collect  the  same  two 
sorts  of  lists  and  was  only  given  lists  arranged  in  sibling  pairs. 
When  I  asked  for  the  names  paired  as  kaishi  and  said  I  had 
been  given  such  lists  in  Mindimbit,  my  informant  said,  "Yes, 
the  Easterners  are  always  talking  about  kaishi;  with  them, 
everything  has  a  kaishi;  but  we  are  not  like  that."  This  eido- 
logical  difference  between  Palimbai  and  Mindimbit  is  in  part 
due  no  doubt  to  historical  accident,  and  it  is  interesting  that 
the  Easterners  are  geographically  nearer  to  the  area  in  which 
extremely  complicated  symmetrical  kinship  systems  have 
been  collected  among  the  Banaro^  and  Mundugumor.^ 

But  in  spite  of  this  difference  in  degree  between  Palimbai 
and  Mindimbit,  both  types  of  dualist  thought  are  well 
developed  throughout  the  latmul  area;  and  both  are  recog- 
nisable in  the  shaping  of  naven.  On  the  one  hand  we  have 
the  emphasis  upon  the  various  direct  identifications  of 
siblings,  of  father  and  son,  and  of  husband  and  wife,  and  on 
the  other  hand,  as  examples  of  the  diagonal  way  of  thinking, 
we  have  the  whole  patterning  of  the  brother-in-law  relation- 
ship and  the  beautiful  reversed  symmetry  in  naven  on  the 
two  sides  of  the  marriage  tie,  whereby  the  father's  sister 
dresses  up  as  a  man  and  identifies  with  the  father,  while  the 
mother's  brother  dresses  up  as  a  woman  and  identifies  with 
the  mother. 

There  are,  of  course,  other  types  of  dualism  besides  the 
two  which  have  been  cultivated  by  the  latmul.  In  our  own 
culture,  we  have  the  dualism  between  Mind  and  Matter  and 
others  based  on  contrast  between  different  levels  of  the  per- 
sonality. In  these  types  of  dualism  the  latmul  have,  so  far 
as  I  know,  not  indulged ;  neither  have  they  emphasised  at  all 
strongly  any  dualism  based  upon  contrasting  ethological 
patterns.  This  is  perhaps  surprising  in  a  culture  in  which 
each  sex  has  its  own  ethos.  Even  the  common  idea  of  "  male  " 
and  "female"  principles  in  nature  is  only  recognisable  as  a 
very  minor  motif  in  the  totemic  dual  division.  The  Sun  is 

^  Thumwald,  Mem.  Amer.  Anth.  Ass.  vol.  in,  1916. 
^  Margaret  Mead,  Sex  and  Temperament. 


The  Eidos  of  lattnul  Culture  241 

personified  as  masculine  and  the  Moon,  which  belongs  to  the 
Mother  moiety,  is  a  woman.  The  Mother  moiety  has  the 
vulva  (not,  so  far  as  I  know,  personified  but  feminine  in 
gender)  as  its  most  important  emblem.  But  this  type  of 
contrast  is  not  carried  far,  and  the  earth  and  sky  are  both 
personified  as  masculine.  In  general,  latmul  thought  has 
ignored  the  sort  of  philosophy  which  finds  ethological  dif- 
ferentiations in  nature.  There  has  never  been  a  Blake  among 
them ;  and  their  myths  of  origin  of  the  world  emphasise  the 
separation  of  sky  and  earth  rather  than  the  coming  together 
of  great  opposites  to  produce  the  world.  Indeed  there  is  so 
little  tendency  to  link  ethological  contrasts  on  to  diagonal 
dualism  that  with  some  hesitation  I  have  classified  the 
identification  of  man  and  wife  among  those  based  upon  direct 
dualism  instead  of  putting  it  in  the  same  class  with  the 
identification  of  brothers-in-law.  In  marriage,  the  contrast 
in  sex  ethos  makes  impossible  the  element  of  symmetrical  com- 
petition which  characterises  co-operation  between  brothers- 
in-law.^ 

Lastly,  before  leaving  the  questions  of  direct  and  diagonal 
thought,  we  may  note  that  primitive  peoples,  as  well  as 
anthropologists,  tend  to  confuse  the  two  methods  of  thinking. 
In  a  typical  classificatory  system,  the  distribution  of  kinship 
terms  is  based  upon  the  identity  between  siblings  of  the  same 
sex.  Thus  the  use  of  identical  kinship  terms  for  two  different 
persons  becomes  typically  a  device  for  stressing  their  direct 
social  identity,  but  the  same  device  may  also  be  used  to  stress 
identity  based  upon  diagonal  thinking.  Thus  among  the  Sulka 

^  This  paragraph  was  written  before  I  had  equated  direct  dualism  with 
complementary  patterns,  and  I  have  left  the  passage  unchanged.  I  had 
found  that  those  with  whom  I  discussed  the  types  of  dualism  tended  to 
confuse  the  geometrical  oppositeness  of  diagonal  dualism  with  the  sort 
of  oppositeness  which  occurs  in  the  contrast  between  Day  and  Night. 
My  concepts  were  at  the  time  rather  vague  but  I  was  sure  that  this  was 
a  misconception,  and  this  passage  is  therefore  devoted  to  showing  that 
oppositeness  of  nature  is  not  implied  in  diagonal  dualism.  It  did  not 
occur  to  me  to  put  the  matter  as  I  now  should  in  positive  terms,  stating 
that  complementary  nature  is  implied  in  direct  dualism. 

The  lack  of  emphasis  in  the  totemic  system  upon  "  male  "  and  "  female  " 
principles  is  probably  correlated  with  the  fact  that  this  system  is  a  re- 
flection of  the  subdivision  of  the  community  into  moieties  and  clans  rather 
than  of  the  differentiation  of  the  sexes. 


242  The  Eidos  of  latmul  Culture 

of  New  Britain,  whom  I  studied  for  some  months  in  1929, 
the  kinship  term  used  for  own  mother's  brother  is  extended 
to  own  mother's  brother's  counterpart  {krus)  in  the  opposite 
moiety,  a  man  whom  the  speaker  would  address  as  "father" 
were  it  not  for  his  diagonal  identification  with  the  mother's 
brother.  In  the  latmul  system  the  same  phenomenon  is 
recognisable  in  the  observance  by  father's  tambinyen  of  the 
taboo  on  eating  food  collected  by  the  son,  a  taboo  which 
chiefly  concerns  the  father. 

Again,  a  confusion  of  direct  and  diagonal  dualism  occurs 
in  the  system  of  totemic  moieties  and  clans.  This  system  is 
thought  of  as  a  series  of  diminishing  dichotomies,  but  it  is 
not  perfectly  clear  whether  these  dichotomies  are  direct  or 
diagonal.  I  have  already  indicated  that  the  relationship  be- 
tween the  moieties  is,  on  the  whole,  diagonal.  But  pairs  of 
related  clans  are  thought  of  as  descended  from  pairs  of  elder 
and  younger  brothers  and  thus  the  relationship  between  them 
is  conceived  of  as  direct ;  and  this  concept  has  been  extended 
to  the  moiety,  so  that  it  is  said  by  the  Sun  people  that  their 
moiety  is  "elder  brother"  of  the  Mother  moiety.^ 

Equally,  instead  of  stating  that  the  direct  dualism  which 
occurs  between  pairs  of  clans  has  been  extended  to  the  relation- 
ship between  the  moieties,  we  may  say  that  the  diagonal 
dualism  between  moieties  has  influenced  the  relationship 
between  the  clans.  The  lists  of  totemic  ancestors  of  the 
various  clans  show  that  a  great  deal  of  imitation  has  occurred. 
Every  clan  has  its  personified  pig,  eagle,  fire,  crocodile,  cere- 
monial house,  shamanic  spirit,  etc. ;  and  it  is  likely  that  the 
sense  of  symmetry  connected  with  diagonal  dualism  has 
guided  the  various  clans  in  their  copying  of  each  other  (i.e. 

^  I  think  I  was  also  told  by  members  of  the  Mother  moiety  that  their 
group  was  "elder  brother"  of  the  Sun  moiety  but  I  have  no  record  of 
this  in  my  notes.  I  probably  took  but  little  notice  of  the  statement  when 
it  was  made,  ignoring  it  as  a  boast.  But  the  matter  is  interesting,  since  if 
difference  in  age,  the  mark  of  complementary  relationship,  is  become  a 
matter  for  symmetrical  rivalry,  this  might  throw  some  light  upon  the 
alternating  system  of  grouping  of  initiatory  grades  into  moieties  with 
symmetrical  rivalry  between  them.  Cf.  also  the  quarrel  between  Kwo- 
shimba  and  Kili-mali  (p.  105)  who  were  related  symmetrically  through 
the  marriage  of  their  children.  Kwoshimba  finally  boasts  that  he  is 
Kili-mali's  "elder  brother". 


The  Eidos  of  latmul  Culture  243 

that  the  relationship  between  clans  though  nominally  comple- 
mentary and  phrased  in  terms  of  "elder"  and  "younger" 
is  to  some  extent  patterned  upon  symmetrical  rivalry). 

There  is  one  peculiarity  of  the  latmul  handling  of  direct 
dualism  which  needs  to  be  specially  emphasised  (and  from 
this,  more  than  from  any  other  detail  I  shall  in  the  Epilogue 
deduce  an  equivalence  between  "direct"  dualism  and  com- 
plementary patterns  of  relationship).  Every  thing  and  every 
person  has  a  sibling  and  the  polysyllabic  names  are  so  ar- 
ranged in  pairs  that  in  each  pair  one  name  is  the  elder  sibling 
of  the  other.  Throughout  the  whole  field  in  which  direct 
dualist  thought  is  recognisable,  it  is  accompanied  by  the  con- 
cept that  one  of  the  units  is  senior  to  the  other.  There  is  no 
such  concept  in  the  case  of  persons  identified  by  diagonal 
dualism  and  it  would  seem  that  such  persons  are  nominally 
equal  in  status  and  always  of  the  same  sex ;  while  those  who 
are  directly  identified  can  never  be  equal  but  must  differ 
either  in  seniority  or  in  sex. 

In  the  case  of  siblings  of  opposite  sex,  there  is  no  emphasis 
on  seniority,  the  diff"erence  in  sex  being  apparently  analogous 
to  difference  in  age.  A  sister,  regardless  of  age,  always 
addresses  her  brother  as  nyamun,  the  term  used  by  a  man  for 
his  elder  brother  only.  But  between  siblings  of  the  same  sex, 
the  terms  nyamun  (elder  sibling  of  same  sex)  and  tshuamho 
(younger  sibling  of  same  sex)  are  always  employed.  In  the 
naming  system,  there  are  pairs  of  masculine  names  and  in 
many  cases  corresponding  pairs  of  feminine  names,  thus: 

Mwaim-nanggura-ndimi  Mwaim-nanggura-ndimi-ndjowa 

(elder  brother)  (elder  sister) 

Temwa-nanggura-ndimi  Temwa-nanggura-ndimi-ndjowa 

(younger  brother)  (younger  sister) 

etc.  (nine  pairs  of  names),    etc.  (seven  pairs  of  names). 

Always  the  seniority  is  stated  in  the  same  sex  pairs  and 
omitted  between  the  sexes. 

In  other  cultures  in  which  direct  dualist  thought  has  been 
at  work,  we  may  find  it  expressed  in  mythological  references 
to  twins.   But  this  type  of  myth  is,  so  far  as  I  know,  absent 


244  '^he  Eidos  of  latmul  Culture 

from  latmul  culture.  It  is  likely  that  the  extreme  horror  and 
intolerance  of  twins  in  some  cultures  are  a  further  develop- 
ment comparable  with  the  latmul  stressing  of  seniority. 
Among  the  latmul,  twins  are  tolerated  and  no  great  interest 
is  taken  in  the  phenomenon  of  twinning.  There  are,  however, 
a  few  myths  which  embroider  upon  the  idea  of  namesakes 
(oiselt).  One  man  meets  another  and  on  asking  his  name  finds 
it  to  be  the  same  as  his  own.  He  then  asks  the  father's  name 
and  finds  (as  the  naming  system  theoretically  demands)  that 
it  is  the  same  as  that  of  his  own  father.  In  mythology  name- 
sakes immediately  strike  up  a  friendship.  But  in  real  life 
they  seem  to  pay  but  little  attention  to  each  other. 

Leaving  now  the  instances  of  paired  identification  or  oppo- 
sition and  turning  to  longer  series,  a  new  pattern  of  thought 
is  recognisable.  In  latmul  culture,  with  its  lack  of  chieftain- 
ship or  hierarchy,  there  are  but  three  contexts  in  which  such 
seriation  occurs,^  and  in  all  of  them  the  first,  third,  fifth,  etc. 
units  are  grouped  together,  and  this  group  is  contrasted  with 
that  consisting  of  the  second,  fourth,  sixth,  etc.  The  three 
contexts  in  which  this  patterning  is  developed  are:  the 
alternation  of  generations,  the  alternation  of  initiatory  grades, 
and  the  alternation  of  siblings. 

A  man's  own  generation,  his  paternal  grandfather's  genera- 
tion, and  his  patrilineal  grandson's  generation,  are  grouped 
together  as  one  mbapma  (literally  "line"),  and  in  contrast  to 
this  his  father's  and  his  son's  generations  are  of  the  opposite 
mbapma.  I  have  already  mentioned  that  theories  of  reincarna- 
tion are  linked  up  with  this  alternation  and  I  need  only  add 
here  that  a  man  may  address  his  father's  father  either  as 
nggwail  (grandfather)  or  as  nyamun  (elder  brother).  This 
identification  of  relatives  with  others  in  analogous  positions 
two  generations  away  on  a  patrilineal  line  runs  through  the 
whole  kinship  system,  so  that  a  man's  son's  wife  is  nyame 
(mother)  and  his  son's  wife's  brother  is  wau  (mother's 
brother). 

The  initiatory  system,  the  second  context  in  which  the 

^  There  are  one  or  tvv'o  minor  instances  of  seriation  without  grading, 
e.g.  the  planting  of  yams  in  rows  and  the  arrangement  of  house  posts. 
In  these  instances  there  is  no  alternation. 


The  Eidos  of  latmul  Culture 


245 


same  sort  of  alternation  occurs,  can  only  be  described  very 
briefly  here. 


Ax  I 

By 

3 

4 

5 

f. 

('•\ 

\7) 

Bx 

3 

■+ 

5 

f, 

(7) 

Figure  4.    Diagram  of  Initiatory  Groups. 

The  two  cross-cutting  moieties  are  marked  A  and  B,  and 
X  and  jy.  Thus  the  resulting  quadrants  are  Ax  and  Ay ;  By  and 
Bx.  Members  of  a  moiety  initiate  youths  of  the  opposite 
moiety.  Thus  the  members  of  Ax  initiate  By,  and  Ay  initiate 
Bx.  Each  quadrant  is  divided  into  three  named  generation 
groups:  I,  3,  5  or  2,  4,  6;  such  that  i  are  the  fathers  of  3, 
who  are  the  fathers  of  5 ;  while  2  are  the  fathers  of  4,  who 
are  the  fathers  of  6. 

The  system  is  so  arranged  that  one  of  the  moieties,  A  or  B, 
is  always  senior  to  the  other.  In  the  Diagram,  A  is  repre- 
sented as  senior  to  B.  Thus,  of  the  generation  groups  in  A, 
each  is  slightly  senior  to  the  corresponding  group  in  B. 
Members  of  Ax  i  initiated  the  members  of  By  2.  Ax  3  are 
now  initiating  By  4.  Ax  5  are  only  just  ready  for  initiation 
but  will  later  be  initiated  by  By  4.  By  6  are  all  small  children. 
Similarly,  Ay  i  initiated  Bx  2,  etc. 

In  course  of  time,  Ax  i  will  die  off  and  a  new  generation  Ax  7 
will  appear.  The  moiety  B  will  then  become  senior  to  A  and 
their  seniority  will  be  recognised  when  they  assert  their  strength 
violently  in  a  brawl.   By  4  will  then  begin  initiating  Ax  5. 

In  this  beautiful  system  the  groups  i,  3,  5  are  united,  and 
opposed  to  2,  4,  6,  which  alternate  with  them  and  are  always 


246  The  Eidos  of  latmiil  Culture 

their  junior  by  about  half  a  generation.  Thus  the  pattern  of 
thought  which  appeared  as  an  alternation  of  generations  in 
the  patrilineal  clan  here  appears  as  an  alternation  of  half- 
generation  groups  which  are  staggered  like  the  spokes  of  a 
wheel. 

Throughout  the  system  each  grade  is  spoken  of  as  the 
"elder  brother"  of  the  grade  next  below.  Thus  i  is  elder 
brother  of  2,  and  2  of  3,  etc.  This  usage  links  up  with  the 
third  context  in  which  the  same  patterning  occurs. 

In  large  families  in  which  there  is  a  long  series  of  brothers, 
the  same  sort  of  alternation  is  expected.  The  language  pos- 
sesses five  terms  (not  numerals)  for  first,  second,  third,  fourth 
and  fifth  child  respectively ;  and  it  is  expected  that,  in  their 
quarrels  over  the  patrimony  in  which  they  share  before  the 
father  dies,  the  first  and  third  brothers  will  join  forces  against 
the  second  and  fourth.  I  asked  about  the  fifth  brother  but 
was  told  "he  would  be  too  little,  he  would  watch  them". 

Lastly  there  is  one  other  context  in  which  I  am  tempted 
to  see  a  development  of  the  same  type  of  pattern,  namely  in 
the  structure  of  flute  music. ^  The  initiatory  flutes  have  no 
lateral  stops  by  which  the  player  might  control  and  vary  the 
pitch.  He  is  therefore  limited  to  the  series  of  natural  har- 
monics provided  by  the  tube.  A  flute  will  give  about  seven 
notes  distributed  over  three  octaves,  but  the  intervals  between 
the  notes  are  too  large  to  admit  of  any  tune  being  produced 
by  a  single  flute.  This  difficulty  is  overcome  by  a  scheme  of 
alternation.  The  flutes  are  always  played  in  pairs  and  in  each 
pair  one  flute  is  one  tone  higher  than  the  other.  In  the  playing, 
as  a  result  of  the  difference  in  pitch  between  the  two  instru- 
ments, the  harmonics  of  the  one  alternate  with  the  harmonics 
of  the  other ;  so  that  the  two  performers,  who  blow  in  turn, 
can  together  produce  simple  tunes.  The  flute  of  lower  pitch 
is  spoken  of  as  the  "elder  brother"  of  the  other,  though  for 
totemic  purposes  the  pair  of  flutes  is  taken  as  a  unit. 

Whether  this  staggering  of  the  harmonics  of  flutes  is  really 

another  expression  of  the  sort  of  thinking  with  which  we  are 

here  concerned  may  perhaps  be  doubted.     But  from  the 

alternation  of  generations,  initiatory  grades,  and  brothers, 

^  The  Eagle,  St  John's  College  Magazine,  1935,  p.  258. 


The  Eidos  of  lattnul  Culture  247 

the  reality  of  the  phenomenon  is  clear  enough.  So  far  as 
I  know,  this  pattern  of  thought  is  not  recognisable  in  the 
shaping  of  naven,  apart  from  such  small  points  as  the  identi- 
fication of  man's  sister's  husband's  father  as  a  laua  (cf. 
distribution  of  pigs,  p.  18).  But  I  have  described  this  alterna- 
tion in  some  detail  because  of  its  general  interest  and  because 
I  believe  that  we  may  see  in  it  a  combination  of  the  two  types 
of  dualist  thinking,  an  intrusion  of  diagonal  thought  into 
contexts  where  direct  identification  and  the  principle  of 
seniority  were  already  at  work.  Thus  between  pairs  of 
brothers  only  these  two  factors  are  concerned,  but  in  longer 
series  we  find  the  principle  of  symmetrical  oppositeness 
breaking  up  the  series. 

In  this  connection  it  is  interesting  to  compare  the  myth 
of  the  Dioscuri,  Castor  and  Pollux,  with  the  alternations 
characteristic  of  latmul  eidos.  Of  these  twins,  Castor  was 
mortal  and  Pollux  was  immortal.  The  mortal  twin  was  finally 
killed  in  battle  and  his  immortal  brother  then  interceded  with 
Zeus  who  granted  him,  as  a  favour,  that  he  should  share  his 
immortality  with  his  brother.  From  that  time,  they  have 
existed  on  alternate  days.^ 

The  resemblance  between  the  alternation  of  the  Heavenly 
Twins  and  that  of  brothers  and  initiatory  grades  among  the 
latmul  is  striking;  but  we  may  cautiously  ask  what,  if  any- 
thing, it  signifies.  In  the  first  place  it  is  obvious  that  an 
eidological  resemblance  between  two  cultures  need  have  no 
significance  from  a  sociological  or  ethological  point  of  view. 
Still  less  will  it  imply  any  historical  connection  between  the 
cultures.  The  resemblance  between  the  two  systems  has  only 
been  demonstrated  within  strictly  defined  limits  and  we  have 
no  right,  when  we  draw  our  conclusions,  to  wander  outside 
these  limits  into  realms  of  history,  sociology,  or  economics. 
We  cannot  rush  on  to  deduce  from  the  eidological  resem- 
blance that  the  relationship  between  brothers  among  the 
Greeks  and  the  latmul  is  alike  in  any  single    pattern   of 

^  Other  versions  of  this  story  occur  and  are  less  comparable  with  the 
alternations  of  latmul  culture.  It  is  said,  for  example,  that  Pollux  was 
permitted  to  spend  his  days  alternately  among  the  gods  and  with  his 
brother  in  Hades. 


248  The  Eidos  of  latmul  Culture 

cultural  behaviour.  It  is  obvious  that  the  same  eidos  might 
be  expressed  in  the  two  cultures  by  means  of  very  different 
forms  of  behaviour. 

The  only  conclusion  then  which  we  are  entitled  to  draw 
is  that  comparable  patterns  of  thought  occur  in  the  Greek 
and  latmul  cultures,  and  we  may  note  that  in  both  cultures 
these  patterns  are  expressed  in  contexts  of  the  relationship 
between  brothers.  Further  than  this  we  cannot  go.  But  if 
the  strict  isolation  of  eidological,  ethological  and  sociological 
points  of  view  permits  us  even  this  much  freedom  in  com- 
paring the  details  of  one  culture  with  those  of  another  I  be- 
lieve that  it  will  be  worth  while.  Hitherto  we  have  had  no 
criteria  of  equivalence  which  would  indicate  either  what  com- 
parisons might  be  permissible  or  what  sort  of  conclusions 
we  might  base  upon  the  resemblances  which  we  found. 

I  suggest  that,  if  we  discipline  ourselves  to  think  rigidly 
in  terms  of  one  sort  of  social  or  cultural  function  at  a  time, 
we  shall  be  free  at  last  to  compare  a  detail  from  one  culture 
with  a  detail  from  another,  knowing  what  we  are  comparing 
and  what  sort  of  deductions  may  be  made  from  the  com- 
parison. It  is  true  that  our  conclusions  will  be  so  limited 
in  scope  as  to  appear  almost  useless,  but  at  least  they  will 
have  one  use,  that  of  suggesting  further  problems.  While 
eidological  conclusions  can  throw  no  light  upon  other  aspects 
of  cultural  behaviour,  they  may  at  least  suggest  problems, 
the  answers  to  which  must  be  sought  in  terms  of  these  other 
aspects.  The  detection  of  similar  eidos  in  two  different  cul- 
tures may  well  stimulate  an  enquiry  into  the  sociology  or 
economics  of  the  peoples  concerned. 

■^  Another  comparative  point  which  will  almost  certainly 
repay  investigation  may  be  seen  in  the  resemblance  between 
the  latmul  kinship  system  and  the  class  systems  of  Australia, 
Ambrym,  etc.  In  all  these  systems  we  may  recognise  both 
diagonal  and  direct  dualist  thought,  and  in  all  of  them  we 
find  an  emphasis  upon  the  alternation  of  generations.  One 
conspicuous  difference,  however,  between  the  Australian 
systems  and  that  of  the  latmul  is  that  the  former  are  closed. 
An  Australian  community  is  divided  into  a  fixed  number  of 
groups,  and  it  is  laid  down  which  group  shall  marry  which 


The  Eidos  of  latmul  Culture  249 

group  and  which  generation  shall  marry  which  generation. 
This  is  not  so  in  the  latmul  system,  and  even  if  the  latmul 
confined  themselves  more  rigidly  to  iai  marriage  they  would 
not  have  a  closed  system  like  those  of  Australia.  Here, 
I  believe  that  we  are  concerned  with  a  rather  important 
motif  of  latmul  thought.  The  natives  see  their  community, 
not  as  a  closed  system,  but  as  an  infinitely  proliferating  and 
ramifying  stock.  A  clan  will  grow  big  and  it  will  subdivide; 
a  village  will  grow  big  and  it  will  send  out  colonies.  The  idea 
that  a  community  is  closed  is  probably  incompatible  with 
this  idea  of  it  as  something  which  continually  divides  and 
sends  out  oflPspring  "like  the  rhizome  of  a  lotus". 

It  is  worth  speculating  about  another  difference  between 
Australian  kinship  and  that  of  the  latmul.  It  would  appear 
from  the  accounts  of  Australian  kinship  that  these  people 
think  of  the  assignment  of  kinship  terms  as  determined  by 
the  system  of  social  groups ;  in  fact,  that  in  Australia  classi- 
ficatory  kinship  is  really  classificatory .  In  contrast  to  this, 
the  latmul  certainly  assign  kinship  terms  rather  on  a  basis. 
of  extension  from  the  family  itself.  We  are  not  concerned 
here  with  the  question  of  how  in  the  dim  past  such  and  such 
a  term  came  to  be  assigned  to  such  and  such  a  relative,  but 
rather  with  the  purely  synchronic  question  of  what  method 
of  thought  the  people  use  now  in  assigning  kinship  terms  to 
their  various  relatives,  and  it  is  clear  that  in  a  classificatory 
system  it  is  possible  to  argue  either  from  the  social  groups 
to  the  family  or  from  the  family  to  the  social  groups.  It  is 
probable  at  least  that  some  peoples  do  one  while  other  peoples 
do  the  other.  In  the  case  of  the  latmul,  I  should  say  that  this 
people  assigns  terms  upon  a  basis  of  extension  rather  than 
of  classification. 

This  is  a  matter  about  which  there  has  been  a  certain 
amount  of  discussion  in  anthropological  circles,  and  it  would 
appear  that  some  anthropologists  are  prone  to  one  way  of 
thinking,  while  others  are  prone  to  the  other.  It  would  be 
important  to  know  whether  some  native  peoples  differ  from 
others  in  this  same  respect.  It  is  likely  that  some  cultures 
promote  the  inductive  methods  of  thinking,  while  others 
promote  the  deductive. 


250  The  Eidos  of  latmul  Culture 

If  there  be  a  possibility  that  the  ways  of  thought  of  anthro- 
pologists occur,  too,  among  primitive  peoples,  what  then  are 
we  to  say  of  the  various  ways  of  thinking  which  have  been 
advocated  in  this  book?  We  may  profitably  at  this  stage  go 
back  over  the  system  which  we  have  constructed  and  consider 
what  it  means  in  terms  of  native  eidos.  We  have  documented 
the  fact  that  ever^'  piece  of  behaviour  is  relevant  (a)  to  society 
as  a  whole,  (b)  to  the  emotions  of  the  individual,  and  {c)  to 
the  thought  of  the  individual,  and  we  have  trained  ourselves 
to  think  separately  in  terms  of  each  one  of  these  forms  of 
relevance.  In  addition  to  these  forms  of  relevance  there  are 
others  which  we  might  have  studied :  notably  we  might  have 
considered  the  whole  culture  as  a  mechanism  oriented  to- 
wards the  production,  distribution  and  consumption  of 
material  objects,  or  again  as  a  mechanism  which  moulds  the 
personalities  of  the  individuals. 

The  suggestion  which  I  now  wish  to  put  forward  is  that 
native  peoples  themselves  are  more  or  less  conscious  of  these 
various  aspects  of  their  behaviour,  and  that  cultures  may 
differ  profoundly  in  the  extent  to  which  one  or  the  other 
aspect  is  emphasised^  in  the  consciousness  of  the  in- 
dividuals. 

If  we  ask  an  individual  why  he  behaves  in  a  particular  way, 
there  are  certainly  five  types  of  answer  which  he  may  give. 
He  may  say : 

1.  "This  behaviour  promotes  the  well-being  of  the  com- 
munity", answering  our  question  in  sociological  terms. 

2.  "This  behaviour  is  customary,  or  traditional,  or  con- 
sistent with  other  behaviour",  answering  our  question  in 
structural  terms. 

3.  "This  behaviour  gives  me  certain  types  of  emotional 
satisfaction";  or  "it  gives  emotional  satisfaction  or  dissatis- 
faction to  other  people,  and  I  enjoy  that",  answering  our 
question  in  ethological  terms. 

*  It  is  doubtful,  perhaps,  to  what  extent  this  sort  of  standardisation 
should  be  regarded  as  either  ethos  or  eidos,  but  I  do  not  think  that  this 
question  need  bother  us  materially  at  present.  It  is  possible  that  some 
other  term  should  be  introduced  for  these  various  sorts  of  standardisation, 
but  the  syndromes  associated  with  them  have  not  yet  been  sufficiently 
investigated  to  justify  the  introduction  of  further  terminology. 


The  Eidos  of  I atmul  Culture  251 

4.  "This  behaviour  will  supply  such  and  such  useful 
objects  ",  answering  our  question  in  economic  terms. 

5.  "  This  behaviour  will  train  my  mind,  or  somebody  else's 
mind",  or  "it  results  from  my  upbringing",  answering  our 
question  in  developmental  terms. 

But,  as  we  have  seen,  at  any  rate  in  the  case  of  the  first 
three  points  of  view,  all  the  sorts  of  relevance  are  actually 
present  in  all  behaviour. 

The  question  which  we  have  to  decide  is,  granted  that  all 
the  aspects  are  present,  Is  the  individual  equally  conscious 
of  them  all  ?  I  would  suggest  that  in  different  cultures  these 
different  aspects  are  stressed  in  varying  degrees,  and  that  a 
profound  difference  appears  in  a  personality  according  as  one 
or  the  other  aspect  of  behaviour  is  chiefly  conscious. 

In  enquiring  into  this  matter,  however,  one  precaution  is 
very  necessary.  The  reader  will  have  noticed  that  a  certain 
effort  of  mental  acrobatics  is  required  to  shift  the  mind  from 
thinking  in  terms  of  one  point  of  view  to  thinking  in  terms 
of  another,  and  further  that  he  himself  thinks  more  readily 
in  terms  of  one  or  another  of  the  points  of  view.  I  myself 
tend,  at  least  in  conscious  levels  of  my  mind,  to  think  most 
readily  in  structural  terms;  and  in  estimating  the  relative 
emphasis  in  latmul  culture  of  these  different  methods  of 
thought  I  may  have  been  influenced  by  this  bias.  It  is 
probable  that  the  questions  which  I  asked  of  my  informants 
were  so  phrased  as  to  indicate  that  I  expected  an  answer  in 
structural  terms. 

We  have  seen  that  in  latmul  culture,  sociological  concepts 
are  usually  expressed  in  symbolic  terms.  The  well-being  and 
fertility  of  the  community  is  ascribed  to  the  spiritual  beings 
called  mbwan  or  kop\  the  fighting  force  of  the  village  is 
symbolised  by  the  eagle ;  and  the  unity  of  the  initiating  grade 
is  represented  by  the  crocodile. 

A  more  interesting  series  of  sociological  and  other  phrasings 
is  provided  by  native  statements  about  the  vengeance  sanc- 
tion (cf.  p.  55).  We  have  seen  that  there  are  various  phrasings 
of  the  matter.  We  have  first  the  phrasing  in  terms  of  ngglambi, 
the  dark  cloud  which  is  seen  over  the  house  of  the  guilty 
man  and  which  may  cause  the  sickness  of  his  relatives.  This 


252  The  Eidos  of  latmul  Culture 

phrasing  with  its  loose  inclusion  of  the  whole  group  of  the 
sinner's  relatives  we  may  regard  as  sociological.  Secondly, 
we  have  the  phrasing  in  terms  of  wagan,  and  this  too  would 
seem  to  be  a  sociological  phrasing,  differing  from  the  first 
in  that  the  group  of  individuals  whose  unity  is  here  symboli- 
cally represented  is,  not  the  group  of  the  relatives  of  the 
offender,  but  the  clan  of  the  offended  man.  It  is  the  wagan 
of  that  clan  who  avenges  the  offence  against  one  of  the  clan 
members.  Thirdly,  we  have  the  phrasing  in  terms  of  indi- 
vidual vengeance  or  lex  talionis.  Here  the  matter  is  stated 
simply  in  terms  of  the  emotions  and  motives  of  individuals 
and  the  symmetry  of  their  behaviour;  while  the  symbolism 
characteristic  of  the  more  sociological  phrasings  has  dis- 
appeared. 

In  all  such  sociological  phrasings  we  may  recognise  the 
development  of  a  system  of  thought  which  fits  in  well  enough 
with  the  theory  of  religion  put  forward  by  Durkheim,  and 
we  may  suspect  that  inasmuch  as  sociological  phrasings  tend 
to  be  symbolic,  this  aspect  of  behaviour  is  not  one  which  is 
emphasised  in  latmul  consciousness. 

It  is  interesting  in  this  connection  to  consider  how  in 
modern  European  "totalitarian"  communities,  whether  they 
be  Fascist  or  Communist,  an  aspect  of  behaviour  which  was 
formerly  almost  ignored  is  now  stressed  by  propaganda  of 
every  sort.  The  individual  is  made  to  see  his  behaviour  more 
and  more  as  relevant  to  the  State.  Conscious  phrasing  of  the 
sociological  motive  in  terms  of  the  State  seems  to  replace 
a  former  symbolic  phrasing  of  the  matter  in  terms  of  a  deity. 

Structural  phrasings,  on  the  other  hand,  seem  to  me  to  be 
much  more  developed  in  latmul  culture.  The  wau  is  articu- 
lately stated  to  be  a  mother,  and  much  of  his  behaviour  is 
seen  by  the  natives  as  consistent  with  this  major  premise. 
A  great  deal  of  cultural  behaviour,  too,  is  seen  by  the  natives 
as  "traditional",  and  this,  I  think,  is  probably  an  expression 
of  a  structural  point  of  view.  The  statement  is  one  which 
involves  no  elaborate  syllogism:  it  amounts  on  examination 
merely  to  saying  that  a  thing  is  done  because  it  has  been  done 
in  the  past.  But  this  is,  after  all,  a  statement  of  structural 
consistency. 


The  Eidos  of  latmul  Culture  253 

Another  feature  of  latmul  culture  which  is  probably  an 
expression  of  structural  thinking  is  the  native  tendency  to 
pithy  cliches.  "  She  was  a  fine  woman" — so  they  married  her 
inside  the  clan.  I  do  not  know  exactly  to  what  extent  phrases 
of  this  sort  are  cliches  or  proverbs,  but  they  were  enunciated 
by  my  informants  with  an  intonation  quite  different  from 
that  which  would  accompany  a  statement  of  the  motives  for 
behaviour. 

Another  example  of  the  latmul  interest  in  the  structural 
aspects  of  behaviour  may  be  seen  in  their  insistence  upon  the 
schematic  qualities  and  the  straightness  of  their  marriage 
system  (p.  9 1 )  and  of  their  lists  of  totemic  names  (pp.  1 27, 128). 

We  may  suspect  that  this  emphasis  upon  the  structural 
viewpoint  is  linked  up  with  a  lack  of  emotional  ease  which 
we  have  seen  to  be  characteristic  of  the  latmul  men  and,  if 
this  correlation  be  correct,  then  there  is  a  possibility  that  a 
structural  view  of  the  world  is  an  important  symptom  in  the 
syndrome  which  Kretschmer  has  called  schizothymia ;  and 
questions  arise  as  to  the  possibility  of  schismogenesis  between 
personalities  which  emphasise  one  of  these  points  of  view  and 
personalities  which  emphasise  another. 

Phrasings  in  terms  of  emotion  are,  on  the  other  hand, 
rather  rarer  among  latmul  men,  and  the  only  two  which  I  can 
recall  are  phrasings  of  unpleasant  emotion — ngglangga  and 
kianta — the  former  a  reference  to  the  emotion  of  a  man  whose 
pride  has  been  hurt,  which  might  perhaps  be  translated  as 
"  pique  ",  the  latter  a  term  for  jealousy.  Both  of  these  terms 
are  occasionally  used  in  stating  the  reasons  for  behaviour, 
but  I  never  heard  any  reference  to  pleasant  emotions  as  the 
cause  of  any  detail  of  behaviour.  The  term  wowia  kugwa, 
"to  be  in  love",  is  used  almost  solely  of  women,  and  it  is 
likely  that  among  the  women  emotional  phrasings  of  reasons 
for  behaviour  are  very  much  more  frequent  than  among  the 
men. 

Economic  phrasings  occur  not  infrequently  in  the  culture. 
But  though  records  are  kept  of  every  bride  price,  and  though 
a  latmul  man  is  very  capable  of  driving  a  hard  bargain,  there 
is  in  this  Papuan  culture  none  of  the  enormous  economic 
emphasis  which  is  so  characteristic  of  the  majority  of  Mela- 


254  The  Eidos  of  latmul  Culture 

nesian  peoples.  Currency  and  trade  and  the  accumulation 
of  wealth  could  never  be  described  as  the  major  pre-occupa- 
tions  of  the  latmul.  It  might  be  worth  enquiring  into  the 
relationship  between  cultural  emphasis  upon  economic  aspects 
of  behaviour  and  the  type  of  personality  called  anal  in  psycho- 
analytic jargon. 

Another  point  of  view  which  a  native  people  may  adopt 
is  the  calendric,  and  here  I  can  only  speak  very  tentatively. 
I  myself  have  so  little  appreciation  of  time  that  I  omitted 
almost  entirely  to  enquire  into  this  aspect  of  latmul  culture. 
My  impression  is,  however,  that  among  the  latmul  there  is 
very  little  emphasis  upon  this  aspect  of  behaviour.  The  most 
important  men,  with  whom  it  rested  to  decide  when  a  given 
ceremony  should  occur,  gave  me  occasionally  rather  inarticu- 
late statements  about  this  periodicity  of  wagan  and  mwai 
ceremonies,  but  though  I  once  or  twice  tried  to  get  this 
periodicity  clear,  I  never  succeeded:  my  informants  seemed 
to  vary  in  their  opinions.  Even  on  the  subject  of  the  lunar 
calendar  informants  were  but  little  interested  and  obviously 
ignorant.  The  difficulty  here  was  that  the  calendar  is  not 
simply  lunar,  but  is  also  controlled  by  the  water  level. 
Nominally  the  year  consists  of  twelve  moons,  of  which  five 
are  moons  of  high  water  and  five  are  moons  of  low  water. 
Between  each  of  these  groups  of  five  moons,  there  is  an  inter- 
mediate moon.  But  inasmuch  as  the  rising  and  falling  of  the 
river  is  very  irregular,  my  informants  were  generally  doubtful 
as  to  what  month  it  was  at  any  given  time.  The  astronomical 
year  of  365  days  contains  of  course  approximately  thirteen 
lunar  months,  so  that  with  their  theory  of  twelve  moons  in 
the  year,  the  latmul  could  never  be  precise  in  the  identifica- 
tion of  the  moons  in  their  calendar. 

In  general  the  occupations  of  the  men  seemed  to  be  but 
little  regulated  by  the  time  of  day,  and  the  performance  of 
even  the  more  important  ceremonies  is  liable  to  be  postponed 
in  the  most  erratic  manner.  Ceremonies  which  are  to  be 
performed  at  night  usually  take  place  when  the  moon  is  full, 
but  they  are  not  infrequently  postponed  from  one  full  moon 
to  another,  and  sometimes  this  postponement  leads  to  a  dance 
being  held  after  the  water  has  risen  and  the  dancing  ground 


The  Eidos  of  latmul  Culture  255 

is  flooded.   In  such  cases  a  platform  is  constructed  on  canoes 
for  the  use  of  the  dancers. 

I  have  stated  that  ethos  and  eidos  are  an  expression  of  the 
affective  and  cognitive  aspects  of  personahty,  but  it  is  worth 
stressing  that  other  aspects  of  the  personahty  may  also  be 
standardised,  and  we  may  here  very  briefly  consider  some 
of  these  aspects : 

1.  Apollonian  and  Dionysian.  Dr  Benedict^  has  described 
two  extreme  poles  of  a  possible  variation  in  personality,  and 
has  shown  that  these  extremes  may  be  standardised  in  culture. 
It  is  not  perfectly  clear,  however,  how  the  syndromes  which 
she  has  called  Apollonian  and  Dionysian  are  related  to  ethos 
and  eidos.  My  own  impression  is  that  in  Apollonian  per- 
sonality we  have  a  standardisation  which  might  occur  with 
many  types  of  organisation  of  the  emotions  and  sentiments. 
But  until  the  phenomena  of  dissociation  have  been  properly 
related  to  other  psychological  phenomena,  and  especially 
until  we  have  some  idea  of  what  we  mean  by  consciousness, 
it  will  not  be  possible  to  define  these  poles  of  variation  more 
precisely. 

2.  Tempo.  It  is  a  common  impression  of  those  who  visit 
foreign  countries  that  the  natives  are  either  faster  or  slower, 
brighter  or  duller  in  their  reactions  than  the  members  of  the 
observer's  own  community.  This  impression  is  no  doubt  due 
to  some  form  of  cultural  standardisation  of  the  personalities 
concerned,  and  should  be  investigated.  We  do  not  know,  for 
example,  whether  reaction  time  can  be  affected  by  culture. 

3.  Perseveration.  In  some  cultures  it  appears  that  the 
individuals  take  great  precautions  to  avoid  receiving  or  giving 
each  other  sudden  emotional  shocks.  Miss  Lindgren  tells 
me  that  among  the  Mongols  a  man  who  comes  with  exciting 
news,  even  though  his  news  be  of  a  nature  to  demand  hasty 
action,  will  not  impart  it  till  some  time  has  elapsed.  It  is 
likely  that  cultures  vary  considerably  in  the  flexibility  which 
they  demand  of  the  individual,  and  it  would  be  important 
to  know  something  of  the  position  of  an  over-flexible  indi- 
vidual in  a  culture  which  expects  a  high  degree  of  persevera- 

^  Patterns  oj  Culture,  1935. 


256  The  Eidos  of  latmul  Culture 

tion,  and  the  fate  of  an  emotionally  stiff  individual  in  a 
culture  which  demands  flexibility. 

4.  Lastly  we  may  remember  that  this  book  has  been 
devoted  solely  to  a  synchronic  view  of  latmul  culture,  and  it 
is  possible  that  individuals  and  cultures  vary  in  the  extent 
to  which  they  see  the  world  either  as  a  product  of  the  past  or 
as  a  mechanism  working  in  the  present. 

At  present  it  is  not  profitable  to  do  more  than  suggest  the 
existence  of  these  other  types  of  standardisation  of  per- 
sonality, but  it  is  possible  that  a  study  of  these  psychological 
phenomena  in  their  cultural  settings  will  contribute  some- 
thing to  our  understanding  of  the  phenomena  themselves; 
and  that  cultural  anthropology  has  important  contributions 
to  make  to  the  subject  of  individual  psychology.  Advances 
in  either  of  these  subjects  will  always  contribute  to  advances 
in  the  other,  and  the  concept  of  cultural  standardisation  of 
personality  is  a  link  between  them. 


CHAPTER  XVI 

Epilogue  ig^6 

THE  writing  of  this  book  has  been  an  experiment,  or 
rather  a  series  of  experiments,  in  methods  of  thinking 
about  anthropological  material,  and  it  remains  to  report 
upon  how  I  was  led  to  carry  out  these  experiments,  to  evaluate 
the  techniques  which  I  devised,  and  to  stress  what  I  regard 
as  my  most  important  results. 

My  field  work  was  scrappy  and  disconnected — perhaps 
more  so  than  that  of  other  anthropologists.  After  all,  we  set 
out  to  do  the  impossible,  to  collect  an  exceedingly  complex 
and  entirely  foreign  culture  in  a  few  months;  and  every 
sincere  anthropologist  when  he  comes  back  to  England  dis- 
covers shocking  gaps  in  his  field  work.  But  my  case  was 
somewhat  worse.  The  average  anthropologist  has  some  de- 
finite interest  in  some  one  aspect  of  the  culture  he  is  studying, 
whether  it  be  in  historical  reconstruction,  in  material  culture, 
in  economics,  or  in  functional  analysis,  and  he  will  at  least 
collect  an  adequate  supply  of  material  for  a  book  permeated 
by  his  particular  point  of  view.  But  I  had  no  such  guiding 
interest  when  I  was  in  the  field ;  I  was  (and  still  am)  sceptical 
of  historical  reconstructions;  I  could  not  (and  still  cannot) 
see  that  orthodox  functional  analysis  was  likely  to  lead  any- 
where; and  lastly  my  own  theoretical  approaches  {Oceania, 
1932,  pp.  484  ff.)  proved  too  vague  to  be  of  any  use  in  the 
field. 

I  did  not  clearly  see  any  reason  why  I  should  enquire  into 
one  matter  rather  than  another.  If  an  informant  told  me  a 
tale  of  sorcery  and  murder,  I  did  not  know  what  question 
to  ask  next — and  this  not  so  much  from  lack  of  training  as 
from  excess  of  scepticism.  In  general  therefore,  apart  from  a 
few  standard  procedures  such  as  the  collection  of  genealogies 
and  kinship  terminology,  I  either  let  my  informants  run  on 
freely  from  subject  to  subject,  or  asked  the  first  question 
which  came  into  my  head.  Occasionally  I  made  an  informant 


258  Epilogue  1936 

go  back  to  some  previous  subject  of  conversation,  but  I 
should  have  found  it  hard  to  give  theoretical  reasons  for  such 
special  attention  to  certain  subjects. 

This  method,  or  lack  of  it,  was  wasteful,  but  it  has  its 
compensations.  I  know,  for  example,  how  great  a  value  the 
natives  set  upon  their  enormous  system  of  totemic  names; 
and  I  know  this,  not  from  some  colourless  statement,  but 
from  the  curious  experience  of  writing  down  thousands, 
literally,  of  names  at  my  informants'  bidding ;  and  I  know  it 
too  from  the  more  bitter  experience  of  finding  my  note-books 
full  of  names  when  I  searched  them  for  information  about 
preferred  types. 

The  arrival  of  a  part  of  the  manuscript  of  Dr  Benedict's 
Patterns  of  Culture,  an  event  already  mentioned  in  the 
Foreword,  together  with  conversations  with  Drs  Fortune  and 
Margaret  Mead,  gave  me  a  vague  clue  to  what  I  wanted  to 
do  in  anthropology,  and  in  the  last  three  months  before  I  left 
New  Guinea  I  tried  to  follow  this  up.  I  realised  then  the 
importance  of  method  in  field  work,  and  endeavoured  in  the 
last  part  of  my  time  to  make  up  for  some  of  the  deficiencies 
in  my  notes  on  the  most  ordinary  subjects.  I  recognised  the 
significance  of  the  contrast  in  ethos  between  the  sexes,  but 
I  did  not  devise  any  special  methods  of  enquiry  which  might 
have  been  relevant  to  an  ethological  approach.  It  was  only 
after  I  had  returned  to  England  that  I  realised  the  importance 
of  observing  the  reactions  of  one  sex  to  the  behaviour  of  the 
other,  or  of  collecting  native  statements  about  preferred 
types.  When  an  informant  told  me  that  Woli-ndambwi  had 
a  big  nose  (cf.  p.  162)  I  wrote  his  statement  down,  but  with 
no  idea  that  this  detail  of  the  culture  might  be  of  any  par- 
ticular interest. 

I  am  here  stressing  the  lack  of  method  in  my  field  work 
in  order  to  satisfy  those  who  may  say  that  I  have  ' '  selected 
my  facts  to  fit  my  theories  ".  Some  selection  has,  of  course, 
occurred  in  the  process  of  creaming  my  note-books,  but  this 
has  been  done  in  England.  The  actual  noting  of  facts  was 
done  at  random  before  I  even  dreamt  of  ethos,  eidos  and 
schismogenesis,  and  I  emphasise  these  considerations  now 
because  they  will  never  be  applicable  to  my  future  work 


Epilogue  1936  259 

when  I  hope  to  select  facts  which  shall  be  relevant  to  my 
problems. 

When  I  came  to  the  task  of  fitting  my  observations  together 
into  a  consecutive  account,  I  was  faced  with  a  mass  of  the 
most  diverse  and  disconnected  material.  I  had  for  example 
a  number  of  stories  of  sorcery  and  retribution,  but  to  not 
one  of  them  was  appended  a  systematic  study  of  the  relevant 
facts  of  kinship,  technique,  emotional  attitude,  etc.  I  had 
collected,  of  course,  not  isolated  facts,  but  facts  in  little 
bunches:  the  facts  in  some  bunches  were  grouped  on  a 
chronological  basis,  and  in  other  bunches  they  were  grouped 
on  a  structural  basis,  and  so  on.  No  one  system  of  organisa- 
tion ran  through  the  material,  but  in  general  my  groups  of 
facts  had  been  put  together  by  my  informants,  so  that  the 
systems  of  grouping  were  based  upon  native  rather  than 
scientific  thought.  Out  of  this  material  I  had  to  construct 
a  picture  of  latmul  culture.  If  in  the  circumstances  I  have 
managed  to  demonstrate  some  coherence  in  the  culture,  this 
achievement  is  the  best  testimony  to  the  usefulness  of  the 
methods  of  analysis. 

An  account  of  the  various  steps  which  have  led  to  the  con- 
struction of  this  book  will  serve  as  a  summary  of  my  sug- 
gestions, and  the  narration  of  the  errors,  into  which  I  fell 
by  the  way,  may  be  of  use  to  any  who  care  to  follow. 

One  detail  of  the  Palimbai  naven  had  struck  me  with  great 
force.  I  had  previously  seen  transvestite  women,  proud  of 
their  male  ornaments,  and  had  even  published  an  account 
of  the  Mindimbit  naven  ceremonies,  but  I  had  not  before  seen 
the  transvestite  wau.  I  had  never  realised  that  he  was  a  figure 
of  fun.  My  whole  mental  picture  of  naven  had  been  wrong, 
and  wrong  because,  though  I  had  been  told  what  was  done, 
I  had  no  idea  of  the  emotional  aspects  of  the  behaviour. 

Though  I  did  not  know  what  it  meant,  I  knew  that  the 
wau's  buffooning  had  altered  my  whole  conception  of  naven, 
and,  if  this  was  so,  the  contrast  between  the  bedraggled 
transvesticism  of  the  men  and  the  proud  wearing  of  homicidal 
ornaments  by  the  women  must  somehow  contain  an  important 
clue  to  latmul  culture.  The  change  in  my  way  of  thinking  had 
arisen  from  the  addition  of  emotional  emphasis  to  what  was 


26o  Epilogue  1936 

originally  a  purely  formal  picture,  and  so  I  came  to  believe 
that  ethos  was  the  thing  that  mattered. 

Much  later,  after  I  had  returned  to  England,  I  found  from 
my  photographs  that  women,  when  they  are  decorated  for 
public  ceremonial,  wear  ornaments  such  as  are  usually  worn 
only  by  men,  and  this  discovery  led  me  to  consider  the 
analogy  of  the  fashionable  horsewomen  and  to  develop  the 
theory  of  latmul  transvesticism  which  I  have  put  forward 
in  Chapter  xiv. 

How  sound  this  theory  may  be  I  can  hardly  judge,  but  it 
is  put  forward  in  all  seriousness,  and  I  have  a  special  affection 
for  it  since,  though  now  the  theory  stands  as  a  rather  unim- 
portant detail  in  the  book,  it  was  from  this  detail  that  the 
whole  synthesis  grew.  At  the  International  Congress  in  the 
summer  of  1934,  I  read  a  paper  to  the  Sociology  Section  on 
my  theory  of  latmul  transvesticism.  In  it,  I  gave  sketches 
of  the  naven  ceremonial  and  of  the  ethos  of  each  sex  in  latmul 
culture.  I  felt  that  I  had  accounted  rather  satisfactorily  for 
the  curious  details  of  naven,  and  I  was  still  sure  that  ethos  was 
the  thing  that  mattered. 

After  the  Congress,  I  started  to  rewrite  this  small  paper  on 
transvesticism,  thinking  that  it  was  worth  publishing,  and 
that  I  could  afterwards  use  the  paper  as  the  skeleton  of  a  book 
on  latmul  culture.  But  the  paper  grew  by  the  addition  first 
of  one  method  of  approach  and  then  of  another,  until  it  has 
become  the  present  book,  and  now  its  purpose  is  no  longer 
to  put  forward  a  theory  of  latmul  transvesticism,  but  to  sug- 
gest methods  of  thinking  about  anthropological  problems. 

The  first  difficulty  that  I  encountered  was  over  the  fact 
that  latmul  men  wear  aprons  and  latmul  women  wear  skirts. 
I  felt  that  my  ethological  theory  was  a  fairly  satisfactory 
description  of  the  ceremonial  transvesticism,  but  I  could  not 
account  for  the  simple  circumstance  that  on  more  ordinary 
occasions  special  clothes  were  worn  by  each  sex;  and  this 
differentiation  was  clearly  necessary  if  transvesticism  was  to 
occur  in  the  culture.  But  the  problem  did  not  worry  me 
seriously,  and  I  dismissed  the  matter.  It  was  a  rule  of  the 
culture,  a  "formulation",  that  individuals  of  different  sex 
should  wear  different  clothes.  Indeed,  I  dismissed  the  matter 


Epilogue  1936  261 

so  effectively  that  I  have  nowhere  in  the  book  stated  the 
obvious  fact  that  the  differentiation  of  clothing  according  to 
sex  is  one  of  the  factors  which  "promote"  transvesticism. 

The  matter  became  more  serious  when  I  asked  why  it 
should  be  the  mother's  brother  who  performs  those  antics 
and  not  the  father.  Again  I  took  refuge  in  the  terms  "  formula- 
tion" and  "structure".  The  kinship  system,  "built  up"  of 
formulations,  was  part  of  the  structure  of  society,  and  the 
naven  was  "built  upon"  this  structure.  It  appeared,  then, 
that  there  was  another  "thing"  called  "structure"  which 
mattered  in  the  culture. 

While  ethos  appeared  to  me  to  consist  of  preferred  types 
and  behaviour  which  was  expressive  of  emotion,  structure 
consisted  of  the  kinship  system  and  other  "formulations". 
If  a  man  scolded  his  wife,  his  behaviour  was  ethos ;  but  if  he 
married  his  father's  sister's  daughter,  it  was  structure.  I  even 
went  so  far  as  to  think  of  the  structure  as  a  network  of  chan- 
nels which  guided  the  ethos,  and  were  shaped  by  it  (cf.  foot- 
note, p.  121).  Looking  back  now,  it  is  almost  incredible  that 
I  ever  thought  along  such  lines  or  used  such  metaphors,  and 
I  find  it  quite  difficult  to  write  of  my  early  theories  without 
caricature. 

It  took  quite  a  long  time  to  get  away  from  these  fallacies, 
and  I  had  advanced  in  other  directions  before  I  escaped. 
I  added  what  I  used  to  call  "pragmatic  function"  to  my  list 
of  the  subdivisions  of  culture.  This  was  a  mixture  of  the 
"satisfaction  of  the  needs  of  individuals"  and  the  "integra- 
tion of  society".  The  confusion  between  these  two  sorts  of 
social  function  was  a  result  of  muddling  two  almost  un- 
related facts :  that  allegiance  is  needed  by  the  wau,  and  that 
it  also  serves  to  integrate  society.  I  still  regarded  ethos, 
structure,  and  pragmatic  function  as  categories  into  which 
culture  could  be  subdivided,  and  I  had  even  started  to  write 
the  last  chapter  of  the  book  before  I  escaped  from  this  morass. 

In  this  last  chapter  I  intended  to  examine  the  inter-relations 
between  these  different  subdivisions  of  culture,  and  I  found 
that  I  had  given  no  clear  criterion  for  discriminating  the 
elements  of  culture  which  I  would  pigeon-hole  as  ethos  from 
those  which  I  would  pigeon-hole  as  structure  or  pragmatic 


262  Epilogue  1936 

function.  I  began  to  doubt  the  validity  of  my  categories, 
and  performed  an  experiment.  I  chose  three  bits  of  culture : 
(a)  a  wau  giving  food  to  a  laua;  a  pragmatic  bit,  (b)  a  man 
scolding  his  wife ;  an  ethological  bit,  and  (c)  a  man  marrying 
his  father's  sister's  daughter;  a  structural  bit.  Then  I  drew 
a  lattice  of  nine  squares  on  a  large  piece  of  paper,  three  rows 
of  squares  with  three  squares  in  each  row.  I  labelled  the 
horizontal  rows  with  my  bits  of  culture  and  the  vertical 
columns  with  my  categories.  Then  I  forced  myself  to  see 
each  bit  as  conceivably  belonging  to  each  category.  I  found 
that  it  could  be  done. 

I  found  that  I  could  think  of  each  bit  of  culture  structurally ; 
I  could  see  it  as  in  accordance  with  a  consistent  set  of  rules^  or 
formulations.  Equally,  I  could  see  each  bit  as  "pragmatic", 
either  as  satisfying  the  needs  of  individuals  or  as  contributing 
to  the  integration  of  society.  Again,  I  could  see  each  bit 
ethologically,  as  an  expression  of  emotion. 

This  experiment  may  seem  puerile,  but  to  me  it  was  very 
important,  and  I  have  recounted  it  at  length  because  there 
may  be  some  among  my  readers  who  tend  to  regard  such 
concepts  as  "structure"  as  concrete  parts  which  "interact" 
in  culture,  and  who  find,  as  I  did,  a  difficulty  in  thinking  of 
these  concepts  as  labels  merely  for  points  of  view  adopted 
either  by  the  scientist  or  by  the  natives.  It  is  instructive  too 
to  perform  the  same  experiment  with  such  concepts  as  eco- 
nomics, kinship  and  land  tenure,  and  even  religion,  language, 
and  "sexual  life"  do  not  stand  too  surely  as  categories  of 
behaviour,  but  tend  to  resolve  themselves  into  labels  for 
points  of  view  from  which  all  behaviour  may  be  seen. 

The  matter  may  be  re-phrased,  either  in  terms  of  points 
of  view  adopted  by  the  scientist,  or  in  terms  of  aspects  of 
the  behaviour.  We  must  expect  to  find  that  every  piece  of 
behaviour  has  its  ethological,  structural  and  sociological 
significance.^ 

^  This  legalistic  phrasing  of  cultural  structure  I  have  since  dropped, 
and  have  substituted  the  term  premises  for  rules  or  formulations,  empha- 
sising the  consistent  nature  of  cultural  structure  and  ignoring  the  question 
of  whether  it  is  enforced,  and  whether  it  is  articulately  formulated  by 
the  natives. 

*  I  do  not  claim  to  have  detected  the  fallacy  of  "misplaced  concrete- 


Epilogue  1936  263 

This  meant  that  I  had  only  to  keep  clearly  before  me  the 
conviction  that  ethos,  structure,  etc.  were  merely  points  of 
view  or  aspects  of  the  culture,  and  to  look  for  each  aspect 
in  every  piece  of  behaviour  and  in  every  native  statement. 
But  this  was  still  difficult.  In  the  first  place,  the  habit  of 
thought  which  attributes  concreteness  to  aspects  of  phe- 
nomena is  one  which  dies  hard.  The  fallacy  which  Whitehead 
has  pilloried  has  been  an  important  principle  or  motif  of 
European  eidos,  certainly  since  the  days  of  Greek  philosophy. 
It  has  taken  me  over  a  year  to  drop  the  habit  even  partially, 
and  I  fear  that  many  passages  in  the  book  may  be  still  more 
or  less  infected  with  it,  in  spite  of  drastic  revision. 

Another  difficulty  was  that  of  keeping  my  points  of  view 
distinct  and  separate.  Constantly  I  wandered  from  the  point 
of  view  which  was  appropriate  in  the  chapter  on  which  I  was 
working  and  found,  for  example,  that  I  had  inserted  a  para- 
graph of  structural  phrasings  into  a  chapter  on  ethos.  One 
instance  of  this  is  worth  narrating:  I  had  been  much  im- 
pressed by  the  tone  of  the  debate  on  the  profanation  of  the 
junior  ceremonial  house  (cf.  p.  100)  and  from  my  enquiries 
about  this  debate  the  fact  had  emerged  that  there  are  no 
serious  sanctions  inside  an  initiatory  grade.  This  I  recognised 
as  a  fact  of  prime  importance,  and  I  made  a  note  that  this 
lack  of  internal  sanctions  should  be  inserted  into  the  chapter 
describing  ethos  in  the  ceremonial  house.  Later  I  wrote 
several  pages  on  the  matter  for  insertion  there,  and  it  was 
not  until  I  re-read  what  I  had  written  that  I  found  that  these 
pages  were  nothing  but  sociology.  They  then  became  the 
nucleus  of  the  second  half  of  the  sociological  chapter. 

This  example  will  serve  to  show  how  misplaced  concrete- 
ness tends  to  confuse  the  separable  aspects.  My  sequence  of 
thought,  though  inarticulate,  had  run  something  like  this: 
the  tone  of  this  debate  is  striking ;  the  debate  is  "  pure  ethos  " ; 
therefore  the  generalisation  which  I  got  from  that  debate  is 
ethos. 

I  was  bothered,  too,  by  the  feeling  that  if  the  natives  them- 

ness"  independently  of  Whitehead,  and  should  perhaps  never  have  suc- 
ceeded in  disentangling  the  matter  if  conversations  with  C.  H.  Wad- 
dington  had  not  planted  the  seeds  of  Whitehead's  philosophy  in  my  mind. 


264  Epilogue  1936 

selves  discriminated,  as  the  natives  of  Western  Europe  do, 
between  economics  and  law,  and  regarded  these  as  subdivisions 
rather  than  as  aspects  of  culture,  I  ought  to  accept  their  view 
of  the  matter  and  adopt  their  categories.  Later,  when  I  was 
working  on  latmul  eidos  it  occurred  to  me  that  we  ought  to 
look  for  a  standardisation  in  the  various  cultures  of  the  world 
of  those  very  methods  of  thought  which  I  had  myself  been 
advocating:  and  when  I  realised  that  people  might  think 
sometimes  structurally  and  sometimes  in  terms  of  economics 
and  sometimes  in  terms  of  ethos  or  sociology,  then  I  under- 
stood that  I  need  worry  no  more  about  native  ideas  of  the 
subdivision  of  culture.  I  was  justified  in  expecting  every 
aspect  to  be  represented  in  every  bit  of  behaviour,  and  the 
fact  that  the  natives  were  conscious  only  of  one  of  these 
aspects  in  a  given  context  was  a  point  which  might  be 
significant  in  an  eidological  examination  of  the  culture. 

For  me,  the  escape  from  the  fallacy  of  misplaced  concrete- 
ness  was  an  enormous  advance,  and  I  hope  that  others  may 
find  that  the  account  of  my  adventures  helps  them  to  see 
their  problems  more  clearly  and  simply.  From  then  on,  the 
construction  of  the  system  of  abstractions  which  I  have  put 
forward  was  a  comparatively  easy  matter.  But  three  other 
steps  in  the  process  may  be  mentioned,  if  only  to  give  honour 
where  honour  is  due : 

The  separation  of  sociology  from  my  jumble  of  "pragmatic 
functions".  This  was  due  to  Professor  Radcliflfe-Brown. 
I  gave  a  paper  in  Chicago  outlining  my  system  of  abstractions, 
and  Professor  Radcliffe-Bro\Mi  observed  that  I  used  the  term 
"structure"  in  a  sense  different  from  his;  that  he  used  it  to 
refer  to  the  structure  of  society,  while  I  was  using  it  for  what 
he  then  proposed  to  call  "  cultural  structure  ".  Following  this 
up,  I  resolved  my  "pragmatic  function"  into  two  separate 
abstractions:  "sociology"  in  the  strict  sense  of  the  term  and 
what  I  have  called  "motivation"  or  the  "expression  of  ethos 
in  behaviour". 

The  isolation  of  eidos  from  ethos.  This  resulted  from  the 
awkwardness  of  my  use  of  the  term  "  logic  "  in  the  definition 
of  cultural  structure.  I  found  that  a  footnote  was  necessary 
to  explain  that  we  must  expect  to  find  diflferent  sorts  of  "  logic  " 


Epilogue  1936  265 

in  different  cultures,  and  in  writing  this  footnote  I  realised 
that  I  ought  to  describe  the  logic  of  latmul  culture  as  fully 
as  I  could,  and  that  this  was  an  important  aspect  of  the 
culture.  "Ethos",  in  my  old  phrasing,  was  subdivisible  then 
into  two  major  abstractions,  ethos  and  eidos,  which  were 
related  to  each  other  as  the  affective  aspects  of  behaviour 
are  related  to  the  cognitive,  I  therefore  dropped  the  old, 
wider  use  of  the  term  "ethos"  to  cover  all  those  pervasive 
characteristics  of  a  culture  which  can  be  ascribed  to  thie 
standardisation  of  the  individuals,  and  now  use  Dr  Benedict's 
term,  configuration y  in  this  wide  sense.  Dr  Benedict  has  agreed 
with  me  in  conversation  that,  in  her  original  use  of  this  term, 
she  intended  it  to  include  the  standardisation  of  many  dif- 
ferent aspects  of  personality,  and  it  is  convenient  to  have  a 
general  term  for  the  combination  of  all  these  aspects  of 
culture. 

It  is  unfortunate  that,  although  the  awkward  word  " logic" 
led  to  the  concept  of  eidos,  the  working  out  of  this  concept 
has  contributed  virtually  nothing  to  mitigate  the  awkwardness 
of  "logic".  This  I  think  is  perhaps  the  weakest  point  in  the 
exposition  of  my  system.  I  am  reasonably  satisfied  that  cul- 
tural structure  is  an  important  and  isolable  aspect  of  behaviour, 
but  the  exact  phrasing  of  the  sort  of  consistency  characteristic 
of  culture  when  seen  from  this  viewpoint  has  still  eluded  me. 

Closely  connected  with  this  difficulty  is  that  of  discrimi- 
nating the  affective  from  the  cognitive  aspects  of  personality. 
I  am  aware  that  psychologists  are  inclined  nowadays  to  look 
askance  at  the  terms  affective  and  cognitive,  but  I  still  hope 
that  if  we  first  grant  that  these  terms  are  aspects  and  not 
categories  of  behaviour,  and  then  go  on  to  compare  the  ethos 
and  eidos  of  various  cultures,  we  may  in  the  end  arrive  at  a 
better  understanding  of  thought  and  emotion. 

Schismogenesis.  This  concept  has  developed  very  slowly 
from  conversations  with  Drs  Margaret  Mead  and  Fortune. 
Dr  Mead  contributed  an  idea  of  the  very  first  importance, 
that  of  complementary  ethos.  But  for  a  long  time  after  I 
returned  to  England  I  still  thought  of  the  phenomena  in 
static  terms,  as  types  of  ethos  or  types  of  personality.  I  allowed 
myself  to  be  influenced  unduly  by  the  conclusions  of  typo- 


266  Epilogue  1936 

logical  psychology.  Gradually  I  realised  that  if  there  existed 
this  complementary  relationship  between  the  ethoses  of  the 
two  sexes  in  latmul  culture,  it  was  evident  from  the  special 
nature  of  this  relationship  that  each  ethos  must  have  some 
formative  or  directive  action  upon  the  other.  From  such 
concrete  phrasings  of  the  matter  the  concept  of  comple- 
mentary schismogenesis  developed. 

Later,  out  of  a  conversation  about  European  politics  with 
Alan  Barlow,  I  developed  the  idea  of  symmetrical  schismo- 
genesis. The  chapter  on  these  processes  is  a  result  of  com- 
bining these  two  lines  of  thought  and  leavening  the  mixture 
with  a  little  Hegelian  dialectic. 

Such  then  have  been  the  steps  which  have  led  me  to  the 
isolation  of  five  major  points  of  view  for  the  study  of  the 
behaviour  of  human  beings  in  society — structural  (and  eido- 
logical),  emotional  (and  ethological),  economic,  developmental 
and  sociological.  I  have  given  no  samples  of  the  developmental 
and  economic  methods,  but  I  have  tried  to  give  a  sketch  of 
latmul  culture  from  each  of  the  other  points  of  view.  The 
result  has  been  that  in  each  chapter,  according  to  the  method 
adopted,  different  parts  of  the  culture  have  been  set  side  by 
side.  In  the  structural  chapter,  the  most  diverse  contexts 
of  the  culture  have  been  set  side  by  side  to  demonstrate 
the  identifications  relevant  to  the  naven  behaviour  of  the 
mother's  brother.  In  the  sociological  chapter  we  have  seen 
the  naven  ceremony  set  beside  the  marriage  system  as  part 
of  a  great  mechanism  for  the  integration  of  the  society.  In  the 
ethological  chapters,  characteristic  expressions  of  latmul 
emotion  have  been  collected  together  from  the  contexts 
of  everyday  life,  head-hunting  and  initiation,  and  from 
this  collection  we  were  able  to  go  on  to  attribute  specific 
emotional  value  to  the  behaviour  of  the  various  relatives  in 
naven.  Finally  in  the  chapter  on  latmul  eidos,  I  have  set 
side  by  side  the  patterns  of  latmul  totemism,  initiatory  grades 
and  even  flute  music  to  give  a  picture  of  the  ways  of  thought 
of  the  latmul.  In  fact,  whatever  our  method,  our  material 
is  the  same  and  includes  the  whole  ordered  diversity  of 
latmul  behaviour. 

But  though  the  material  for  our  studies  is  the  same  what- 


Epilogue  1936  267 

ever  aspect  of  the  culture  we  intend  to  study,  it  is  by  no  means 
immaterial  in  what  order  we  investigate  the  different  aspects. 
In  attributing  emotional  value  to  the  details  of  naven  be- 
haviour, I  have  insisted  that  any  such  attribution  is  dangerous 
until  the  ethos  of  the  whole  culture  has  been  studied;  and 
therefore  the  general  picture  of  latmul  ethos  necessarily  pre- 
cedes the  chapter  in  which  I  have  tried  to  analyse  the 
emotions  of  the  various  relatives  in  naven — and  even  with 
this  precaution,  this  chapter  is  perhaps  the  least  satisfactory 
in  the  book. 

In  the  study  of  the  logic  of  latmul  culture  I  have  adopted 
the  opposite  procedure.  I  first  outlined  the  premises  upon 
which  the  naven  ceremony  is  based  and  only  later  gave  a 
general  picture  of  latmul  eidos.  The  question  must  therefore 
be  asked :  why,  if  it  is  not  justifiable  to  guess  at  motives  before 
the  whole  ethos  has  been  studied,  is  it  permissible  to  attribute 
behaviour  to  specific  premises  before  the  eidos  of  the  culture 
has  been  sketched? 

This  question  is,  I  suspect,  one  of  considerable  importance, 
and  its  answer  may  well  reflect  upon  the  relation  between 
the  affective  and  cognitive  aspects  of  the  personality.  At 
present  I  can  only  suggest  two  possible  answers:  the  order 
of  procedure  which  I  have  been  forced  to  adopt  may  be  a 
result  either  of  my  own  psychology  or  of  that  of  my  informants. 
I  have  stated  that  in  my  own  conscious  mental  processes, 
the  structural  and  logical  aspects  of  behaviour  are  more 
clearly  seen  than  the  emotional,  and  I  believe  that  the  same 
is  true  of  the  latmul  men.  If  this  observation  be  correct, 
then  in  attributing  emotional  value  to  details  of  behaviour 
I  was  speculating  about  the  more  unconscious  processes  and 
therefore  needed  the  supporting  material  which  could  only 
be  provided  by  a  preliminary  study  of  latmul  ethos.  But  in 
the  case  of  the  logical  aspects  of  behaviour  I  could  hope  for 
some  degree  of  articulateness  in  my  informants  and  could 
proceed  directly  to  the  statement  of  consistency  between 
different  details  of  behaviour. 

Moreover,  a  greater  articulateness  about  the  structural 
aspects  of  behaviour  is  characteristic  not  only  of  the  natives 
but  also  of  Europeans.  The  English  language — and  perhaps 


268  Epilogue  1936 

all  languages — is  not  adapted  to  the  critical  description  of 
emotion.  While  a  syllogism  can  be  stated  in  stark  simplicity,  it 
is  almost  impossible  to  make  any  simple  statement  about  emo- 
tions. The  best  we  can  do  is  to  sketch  an  ethos  or  a  personality 
in  a  diffuse  journalistic  or  artistic  style  and  then  sum  up  that 
sketch  in  a  few  words  whose  significance  is  fixed  by  the 
sketch.  Thus  after  giving  a  sketch  of  latmul  ethos,  I  labelled 
that  ethos  as  "  proud  "  and  "  schizothyme  ".  These  words  are 
capable  of  a  hundred  different  meanings,  but  the  sense  in 
which  I  have  used  them  is  defined  to  some  extent  by  my 
sketch  of  latmul  ethos.  Failing  such  a  preliminary  sketch, 
our  attribution  of  emotional  value  to  behaviour  can  only  be 
guided  by  general  and  probably  fallacious  assumptions  about 
human  nature,  assumptions  which  are  so  general  that  they 
blur  all  differentiation  of  sex,  temperament  and  culture. 

In  this  final  chapter  we  are  trying,  prematurely  perhaps, 
to  get  some  general  view  of  the  different  methods  of  approach 
which  have  been  sampled  in  the  course  of  the  book,  and  one 
point  is  of  special  interest.  If,  whatever  method  we  adopt, 
our  material  is  the  same  and  the  only  difference  between  one 
method  and  another  lies  in  the  arrangement  of  the  material, 
then  it  is  relevant  to  enquire  whether  any  subdivision  of  the 
culture  or  any  other  systematic  phenomenon  has  proved  so 
fundamental  as  to  survive  these  re-shufflings  of  the  details 
of  behaviour  and  to  reappear  in  recognisable  form  in  each 
of  our  pictures. 

The  orthodox  "subdivision"  of  culture  into  such  institu- 
tions as  marriage,  kinship,  initiation,  religion,  etc,  has  entirely 
disappeared ;  but  one  phenomenon  has  persisted  through  all 
the  re-shufflings  and  has  been  described  in  different  terms 
in  each  picture.  In  the  chapter  on  ethological  contrast  and 
competition,  actually  the  last  chapter  to  be  written,  I  stated 
that  though  the  concept  of  schismogenesis  had  developed  out 
of  the  study  of  ethos,  "we  should  be  prepared  to  study 
schismogenesis  from  all  the  points  of  view — structural,  etho- 
logical and  sociological — which  I  have  advocated".  When  I 
wrote  this  sentence  I  intended  to  postpone  the  study  of  the 
sociological  and  structural  aspects  of  schismogenesis  to  some 
future  work.   I  was  aware  that  the  Marxian  historians  had 


Epilogue  1936  269 

occupied  themselves  with  the  economic  aspects  of  the  process 
and  I  had  deduced  from  this  that  we  must  expect  structural 
and  sociological  aspects,  but  I  had  no  idea  what  the  process 
might  look  like  from  these  points  of  view.  It  was  therefore 
a  surprise  to  me  to  find,  on  re-reading  my  manuscript,  that 
I  had  actually  devoted  several  pages  of  my  chapters  on  eidos 
and  sociology  to  the  two  forms  of  schismogenesis ;  and  the  sur- 
prise was  the  greater  inasmuch  as  the  concept  of  symmetrical 
schismogenesis  had  been  a  very  late  addition  to  my  theory. 

The  last  pages  of  the  chapter  on  sociology  (pp.  106  et  seq.) 
are  devoted  to  a  discussion  of  the  various  patterns  of  fission 
which  occur  in  human  societies.  I  here  stressed  the  contrast 
between  the  fission  of  latmul  communities,  which  gives  rise 
to  daughter  groups  with  cultural  norms  similar  to  those  of 
the  parent  society,  and  the  fission  of  European  communities 
which  so  often  produces  daughter  groups  with  cultural  norms 
divergent  from  those  of  the  parent.  I  noted  also  that  the 
production  of  daughter  groups  with  similar  norms  was  asso- 
ciated with  a  peripherally  oriented  system  of  sanctions,  where- 
as the  production  of  divergent  daughter  groups  was  associated 
with  a  centripetal  or  hierarchic  system. 

These  two  types  of  fission  are  simply  describable  in  terms 
of  symmetrical  and  complementary  schismogenesis.  I  stated 
that  in  the  integration  of  latmul  society  the  weakest  links  are 
those  between  affinal  relatives  and  that  it  is  these  links  which 
are  broken  when  the  community  divides,  while  the  links  based 
upon  patrilineal  relationships  persist.  The  "  plane  of  cleavage  " 
of  this  society  is  parallel  to  the  patrilineal  but  cuts  across  the 
matrilineal  and  affinal  bonds. 

We  have  seen  that  in  this  society  the  relationships  between 
father  and  son  and  between  elder  brother  and  younger  brother 
are  complementary,  while  that  between  brothers-in-law  is 
symmetrical.  Thus  it  is  the  symmetrical  relationships  which 
are  chiefly  broken  in  fission,  and  we  may  reasonably  suspect 
that  the  fission  itself  is  in  part  a  result  of  symmetrical  schismo- 
genesis. 

We  found  that  one  of  the  sociological  functions  of  naven 
is  the  strengthening  of  these  affinal  links,  but  it  was  not  per- 
fectly clear  (cf.  p.  96)  why  for  this  purpose  it  should  be 


270  Epilogue  1936 

relationship  between  classificatory  wau  and  laua  which  is 
stressed  rather  than  that  between  classificatory  brothers-in- 
law.  It  would  seem  to  the  European  mind  to  be  the  more 
natural  to  stress  the  latter  relationship.  But  in  the  light  of 
the  schismogenic  position  it  is  reasonable  to  suppose  that 
the  wau-laua  relationship  with  its  accompanying  difference 
in  age  between  the  persons  is  more  consistently  comple- 
mentary in  its  behaviour  patterns  and  therefore  more  adapted 
to  function  as  a  bond  in  a  society  which  seemingly  has  some 
difficulty  in  controlling  symmetrical  schismogenesis.  More- 
over, it  is  worth  noting  that  the  actual  behaviour  of  the  wau 
in  naven  may  be  described  as  an  insistence  upon  the  comple- 
mentary aspects  of  his  relationship  with  his  laua,  at  the 
expense  of  the  symmetrical  aspects.  He  stresses  his  position 
as  "mother"  and  "wife"  of  his  laua,  thus  denying  the  sym- 
metrical aspect  of  the  relationship  connected  with  his  position 
as  "  brother-in-law  ". 

It  is  possible  that  this  insistence  on  the  complementary 
patterns  in  the  wau-laua  relationship  is  a  case  of  the  control 
of  a  symmetrical  schismogenesis  by  admixture  of  comple- 
mentary patterns  of  behaviour  (cf.  p.  193).  But,  since  the 
final  fission  of  the  community  still  follows  a  plane  of  cleavage 
which  cuts  across  these  matrilineal  and  affinal  bonds,  we  must 
suppose  that  this  attempt  to  control  the  symmetrical  schismo- 
genesis is  not  entirely  successful. 

In  contrast  to  the  fission  of  latmul  communities  by  sym- 
metrical schismogenesis,  we  find  that,  in  European  commu- 
nities, the  daughter  groups  generally  separate  from  the  parent 
in  revolt  against  the  elaborate  centripetal  hierarchies — legal, 
religious  and  military — which  are  a  characteristic  feature  of 
the  integration  of  these  communities.  These  hierarchies  are  all 
of  them  built  up  of  series  of  complementary  relationships,  and 
it  is  therefore  reasonable  to  suppose  that  the  fissions  are  either 
wholly  or  in  part  a  result  of  complementary  schismogenesis. 

More  than  this  we  cannot  at  present  say;  and  it  is  likely 
that  in  attributing  the  fission  of  latmul  communities  to 
symmetrical  schismogenesis  and  that  of  European  commu- 
nities to  complementary  schismogenesis  I  have  been  more 
sweeping  in  my  generalisations  than  is  strictly  justified.  These 


Epilogue  1936  271 

sweeping  statements  will,  however,  serve  as  sample  of  the 
generalisations  which  might  be  carefully  documented  by  a 
study  of  schismogenesis  in  its  sociological  aspects. 

While  in  the  sociological  chapter,  schismogenesis  appeared 
in  the  guise  of  a  classification  of  types  of  social  fission,  the 
same  process  was  represented  in  the  chapter  on  latmul  eidos 
as  a  classification  of  types  of  dualism  (pp.  237-247).  I  was 
dissatisfied  with  my  phrasing  of  "direct"  and  "diagonal" 
dualism  and  had  found  considerable  difficulty  in  explaining 
these  concepts  to  others;  but  I  did  not  see  my  way  to  a 
re-phrasing  of  the  matter  until  I  realised  that  for  the  terms 
"direct"  and  "diagonal"  I  could  substitute  complementary 
and  symmetrical  in  the  senses  in  which  these  terms  had  been 
used  in  the  description  of  schismogenic  relationships. 

This  equivalence  is  clear  from  a  consideration  of  the  list 
of  relationships  which  I  have  classed  respectively  as  "  direct" 
and  "diagonal".  The  "direct"  relationships  are  those  be- 
tween elder  brother  and  younger  brother,  brother  and  sister, 
and  man  and  wife ;  and  all  of  these  are  typical  complementary 
relationships.  The  "  diagonal "  relationships  are  those  between 
brothers-in-law  (especially  brothers-in-law  who  have  ex- 
changed sisters)  and  between  kaishi  (counterparts  who  have 
exchanged  presentations  of  valuables,  or  persons  whose 
children  have  inter-married).  Both  of  these  relationships  are 
symmetrical. 

In  the  "direct"  relationships  the  importance  of  the  com- 
plementary patterning  is  evident  from  the  insistence  in  the 
lists  of  totemic  names  on  difference  of  age  between  siblings 
of  the  same  sex.  In  the  case  of  siblings  of  opposite  sex,  it  is 
significant  that  there  is  no  suggestion  of  a  difference  in  age, 
"the  difference  in  sex  being  apparently  analogous  to  dif- 
ference in  age"  (p.  243).  We  may  now  guess  at  the  nature 
of  this  analogy :  these  two  sorts  of  difference  are  alike  in  that 
each  introduces  a  basis  for  complementary  patterns  into  the 
relationship. 

In  the  relationships  between  pairs  of  clans  and  between 
the  two  totemic  moieties  we  observed  that  both  forms  of 
dualism  are  recognisable.  While  the  clans  are  nominally 
grouped  together  in  pairs  as  "elder"  and  "younger",  the 


272  Epilogue  1936 

members  of  the  clans  also  devote  a  great  deal  of  energy  to 
symmetrical  competition  in  pride  of  ancestry  and,  as  a  result, 
the  types  of  dualism  are  tangled  together.  We  may  suggest, 
very  tentatively,  that  the  statement  of  elder  and  younger 
brotherhood  between  related  clans  perhaps  mitigates  the 
symmetrical  competition  between  them.  But  the  matter  is 
not  clear  and  needs  more  study  of  the  actual  behaviour; 
especially  we  need  to  know  in  what  contexts  the  members  of 
a  pair  of  clans  insist  upon  their  brotherhood  and  in  what 
contexts  they  compete  (cf.  footnote,  p.  242).  Without  more 
complete  information  we  must  disregard  these  groupings  as 
sources  of  evidence  for  the  equivalence  between  the  types  of 
dualism  and  the  types  of  schismogenic  relationship. 

This  equivalence  is,  however,  clearly  demonstrable  in  the 
elaborate  systems  of  alternation  of  initiatory  grades  and  of 
brothers.  I  described  this  alternation  as  due  to  "a  combina- 
tion of  the  two  types  of  dualist  thinking,  an  intrusion  of 
diagonal  thought  into  contexts  where  direct  identification 
and  the  principle  of  seniority  were  already  at  work  "  (p.  247). 
In  terms  of  schismogenic  patterns  we  may  now  re-state  this 
combination  of  the  two  forms  of  dualism.  The  relationship 
between  any  two  consecutive  grades  is  complementary  and  is 
marked  by  such  steep  contrasts  as  occur  in  initiation,  where 
the  novice  is  the  "wife  "  of  the  initiators.  But  the  relationship 
between  the  initiatory  moieties  is  symmetrical.  Rivalry  and 
combat  are  culturally  expected  between  the  two  groups  and, 
in  fact,  this  relationship  provided  the  most  clearly  developed 
of  our  examples  of  symmetrical  schismogenesis. 

As  to  the  mechanism  which  determines  this  building  up 
of  alternating  series,  I  have  assumed  (p.  244)  that  the  alter- 
nating pattern  is  itself  a  motif  of  latmul  eidos  and  that  there- 
fore this  pattern  reappears  in  whatever  contexts  are  suitable. 
But  it  is  not  clear  that  this  assumption  is  justified  or  necessary. 
It  is  possible  that  not  the  whole  pattern  but  only  the  two  types 
of  dualism  are  carried  in  the  eidos.  Thinking  for  a  moment 
in  diachronic  terms,  we  can  see  that  it  is  unlikely  that  any 
of  these  complex  alternating  systems  has  been  introduced 
entire  by  culture  contact.  We  therefore  have  to  consider  how 
the  first  of  these  systems  can  have  developed ;  only  after  this 


Epilogue  1936  273 

will  it  be  permissible  to  think  of  the  alternating  pattern  as 
"carried  by  the  eidos",  i.e.  to  suppose  that  the  other  alter- 
nating systems  are  built  up  on  the  model  of  the  first.  More- 
over, I  know  of  no  native  terminology  to  describe  alternation, 
and  in  the  absence  of  such  terminology  it  is  likely  that  the 
alternating  systems  are  formed  afresh  in  each  context.  It  is 
possible  that  the  only  standardisation  provided  by  the  eidos 
is  that  of  the  two  types  of  dualism.  These  the  native  language 
expresses  adequately  by  means  of  kinship  terminology. 

If  this  be  so,  then  we  must  see  the  development  of  alter- 
nating systems  in  latmul  culture  and  their  absence  in  our 
own  culture  as  a  function  of  the  fact  that  among  the  latmul 
both  complementary  and  symmetrical  patterns  are  thought 
of  in  dual  terms,  while  in  Europe  though  we  regard  comple- 
mentary patterns  either  as  dual  or  as  arranged  in  hierarchies, 
we  do  not  think  of  patterns  of  rivalry  and  competition  as 
necessarily  dual.  Rivalry  and  competition  in  our  communities 
are  thought  of  as  occurring  between  any  number  of  persons 
with  no  presumption  that  the  resulting  system  will  be  pat- 
terned upon  any  sort  of  bilateral  symmetry.  Only  if  both 
types  of  relationship  are  habitually  thought  of  in  dual  terms 
is  it  likely  that  alternating  hierarchies  of  the  latmul  type 
will  be  developed. 

Another  line  of  enquiry  may  be  mentioned,  although  no 
answer  can  be  given  to  the  questions  which  it  raises.  As  in 
the  case  of  contrasting  ethos  we  were  led  to  ask  how  the 
appropriate  system  of  emotional  attitudes  is  inculcated  in  each 
individual,  so  here  we  may  ask  how  the  two  methods  of  dualist 
thought  are  inculcated.  To  this  question  I  will  only  reply 
with  the  awkwardly  phrased  suggestion  that  such  habits  of 
thought  are  a  result  of  observation  of,  and  conversation  about, 
patterns  of  relationship. 

But  the  question  itself  is  of  some  academic  interest  in  the 
present  context  because  it  illustrates  perhaps  better  than  any 
other  the  difference  between  ethos  and  eidos;  and  therefore 
the  examination  of  its  terms  may  lead  to  a  proper  formulation 
of  the  difference  between  affective  and  cognitive  aspects  of 
personality. 

In  the  ethological  problem  we   had  to  decide  how  an 


274  Epilogue  1936 

individual  is  moulded  so  that  he  himself  adopts  a  certain  ethos, 
a  certain  system  of  emotional  attitudes  in  his  contacts  with 
other  people  and  the  outside  world.  In  behaviourist  termino- 
logy, we  had  to  decide  how  a  certain  system  of  responses  is 
conditioned.  But  in  the  present  problem  we  have  to  decide 
how  the  individual  is  trained  to  discriminate  and  classify 
relationships  between  persons  and  between  groups.  Thus  we 
are  here  concerned,  not  with  systems  of  responses,  but  with 
the  system  of  classification  which  the  individual  uses  upon 
the  stimuli. 

We  may  therefore  say  that  every  series  of  events  of  the 
type  which  we  may  call  a  stimulus-response  has  both  affective 
and  cognitive  aspects;  and  the  aspect  which  we  shall  study 
is  determined  by  the  way  in  which  we  work.  If  we  take  a 
given  stimulus-response  and  consider  all  the  details  of  the 
response  and  how  they  are  linked  together,  and  if  we  then  go 
on  to  analyse  the  composition  of  other  sets  of  linked  details 
of  response,  we  shall  be  studying  the  organisation  of  responses 
and  shall  arrive  at  a  picture  of  the  affective  aspects  of  the 
personality.  If,  on  the  other  hand,  we  start  from  a  given 
stimulus-response  and  compare  this  with  other  similar  series 
of  events  to  obtain  first  a  list  of  stimuli  which  evoke  a  par- 
ticular detail  of  response  and  then  a  classification  of  stimuli 
according  to  the  responses  which  they  evoke,  we  shall  finally 
arrive  at  a  picture  of  the  cognitive  aspects  of  the  personality. 

In  these  procedures,  it  may  be  noted  that  strictly  speaking 
we  arrive  not  at  pictures  of  the  individual  but  at  pictures  of 
the  events  in  which  the  individual  is  involved.  This  incon- 
sistency disappears  when  we  realise  that  the  term  "per- 
sonality" refers  not  to  the  isolated  individual  but  to  the 
individual  in  the  world. 

This  definition  of  the  terms  "affective"  and  "cognitive" 
may  be  expressed  diagramipatically  (Figs.  5  and  6).  When  we 
find  that  an  individual  responds  with  reaction.  A,  to  the 
stimuli  I,  3  and  5,  we  may  say  that  he  discriminates  these 
stimuli  from  2,  4  and  6,  and  classifies  them  together  on  the 
basis  of  some  quality  which  they  possess  in  common;  and 
we  might  go  on  from  this  to  a  more  elaborate  study  of  the 
whole  system  of  discriminations  and  classifications  which  are 


Epilogue  1936  275 

characteristic  of  that  individual.    Such  a  study  would  be  a 
study  of  the  cognitive  aspects  of  his  personality. 

At  an  early  stage  in  this  book  (p.  25)  I  spoke  of  "  syllogisms  " 
and  "logic"  in  cultural  structure;  and  it  is  of  some  interest 
to  consider  how  this  would  have  to  be  re-phrased  if  future 
work  should  show  this  definition  of  cognitive  aspects  of  per- 
sonality to  be  valuable.    Briefly  we  should  have  to  phrase 

POSSIBLE 
REACTIONS 


Fig.  5.    The  study  of  cognitive 
aspects  of  personality. 


F  6 

Fig.  6.    The  study  of  affective 
aspects  of  personaHty. 


the  matter  not  in  terms  of  logic  but  in  terms  of  classifications 
of  behaviour.  Instead  of  saying:  "The  individual  exhibits 
reaction,  A,  in  response  to  stimuli  i  and  3 ;  stimulus  5  re- 
sembles (or  is  identified  with)  i  and  3 ;  therefore  the  indi- 
vidual exhibits  reaction  A  in  response  to  stimulus  5  ",  we 
should  rather  say :  "  Stimuli  i ,  3  and  5  are  classified  together  ". 
From  this  we  should  proceed  directly  to  the  study  of  the 
system  of  classification,  the  eidos  of  the  culture. 

On  the  affective  side  the  matter  is  not  quite  so  simple. 
If  we  find  that  the  individual  reacts  to  stimulus,  i ,  with  the 
simultaneous  reactions.  A,  C  and  E,  we  may  classify  these 
reactions  together  as  a  syndrome  of  symptoms  of  some 
emotion.  From  this  we  may  go  on  to  a  classification  of  all 
the  series  of  linked  responses  exhibited  by  the  individual 
and  so  get  a  general  description  of  the  affective  sides  of  his 
personality.  But  there  are  difficulties  in  this  view  and  it  is 
possible  that  there  are  other  mechanisms  which  link  responses 


276  Epilogue  1936 

together,  besides  those  systems  which  we  may  describe  as 
"  emotions  ".  For  the  present,  however,  I  know  of  no  phrasing 
of  the  matter  which  so  nearly  translates  the  term"  emotion  " 
into  behaviouristic  language. 

Such  a  phrasing  of  affect  would  I  think  have  had  but  little 
effect  upon  my  general  statement  of  latmul  ethos.   It  is  true 
that  the  description  which  I  gave  of  latmul  ethos  is  very  far 
from  a  systematic  account  of  sets  of  linked  responses,  but  this 
shortcoming  is  due,  not  to  a  weakness  in  my  theoretical 
concepts,  but  to  the  merely  practical  difficulty  of  describing 
human  behaviour  in  a  critical  and  comprehensive  manner. 
When  I  stated  that  the  "tone"  of  the  men's  behaviour  in  the 
initiation  ceremonies  was  expressive  of  harshness  and  irre- 
sponsibility rather  than  of  asceticism,  I  meant  that  the  actions 
i  performed  by  them,  the  washing  of  the  novice,  etc.,  were 
[accompanied  by  other  details  of  behaviour  so  that  the  whole 
picture  was  one  of  harshness.    Until  we  devise  techniques 
ifor  the  proper  recording  and  analysis  of  human  posture, 
'gesture,  intonation,  laughter,  etc.  we  shall  have  to  be  content 
iwith  journalistic  sketches  of  the  "tone"  of  behaviour. 

Some  modification  would,  however,  have  been  necessary 
in  the  chapter  entitled  "  The  Expression  of  Ethos  in  Naven  ". 
The  term  "  motivation  "  would  have  disappeared  entirely,  but 
it  is  not  quite  clear  what  would  have  taken  its  place.  I  think 
that  ideally  the  chapter  should  have  consisted  of  an  analysis 
of  all  the  syndromes  of  response  exhibited  by  the  wau.  These 
syndromes  would  then  have  been  compared  and  equated 
with  syndromes  of  response  collected  in  the  other  contexts 
of  the  culture,  and  finally  the  chapter  would  have  developed 
into  a  synthesis  of  affective  and  cognitive  studies  to  show 
that  the  stimuli  of  the  naven  situation  were  such  as  might 
be  expected  to  evoke  these  syndromes. 

From  this  contemplation  of  the  elaborate  analysis  which 
ideally  ought  to  have  been  done,  we  may  turn  to  what  actually 
was  done  and  examine  more  carefully  the  differences  between 
the  various  points  of  view  which  I  adopted  in  the  various 
chapters.  We  have  seen  that  schismogenesis  is  recognisable 
in  three  different  aspects  of  the  culture,  in  ethos,  in  eidos 
.  and  in  sociology ;   and  we  may  profitably  consider  what 


Epilogue  1936  277 

distortions  were  introduced  into  the  picture  of  this  dia- 
chronic  process  according  to  the  various  methods  of  its 
delineation. 

In  the  jfirst  place,  it  is  certain  that  all  the  representations 
of  schismogenesis  should  have  had  one  failing  in  common. 
They  should  have  given  us  no  indication  that  schismogenesis 
is  a  diachronic  process,  a  process  of  change.  Each  method 
was  nominally  strictly  synchronic  and  each  should  have  given 
only  a  synchronic  "section"  of  schismogenesis  in  which  the 
process  of  change  would  have  been  seen  as  stationary  in  a 
single  instant  of  time. 

Actually,  the  sociological  view  of  schismogenesis,  owing 
to  a  lack  of  stringency  in  my  exposition,  showed  us  the  two 
forms  of  social  fission,  and  these  must  I  think  be  regarded 
as  diachronic  processes.  Strictly  in  the  sociological  chapter 
I  should  have  confined  myself  to  discriminating  between 
peripheral  and  centripetal  systems  of  integration  and  should 
not  have  wandered  from  this  classification  to  consider  the 
results  of  fission  in  the  two  cases.  It  is  conceivable  that  even 
from  this  strictly  limited  classification  we  might  have  arrived 
at  the  statement  that  peripheral  integration  is  chiefly  depen- 
dent upon  symmetrical  links  between  individuals,  while 
centripetal  integration  is  chiefly  dependent  upon  comple- 
mentary links.  But  I  did  not  grasp  this  at  this  stage,  probably 
because  my  attention  was  fixed  too  rigidly  upon  the  major 
unit,  the  society  as  a  whole. 

In  the  description  of  latmul  ethos,  however,  the  picture 
which  we  drew  of  schismogenesis  was  strictly  synchronic. 
We  saw  ethological  contrast  as  statically  existing  in  the  cul- 
ture.^  Indeed  from  the  labelling  of  the  ethological  contrast, 
first  as  "  complementary  "  and  then  in  terms  of  Kretschmer's 
typology,  to  the  final  realisation  that  each  ethos  must  have  a 
formative  effect  on  the  other,  was  a  very  long  step.  The  dia- 
chronic aspects  of  schismogenesis  did  not  come  into  view 
until  I  attacked  the  question  of  the  processes  of  moulding 
of  individuals  into  the  complementary  patterns,  until,  in  fact, 

^  The  analysis  of  the  schismogenic  processes  in  initiation  was  a  late 
insertion  into  the  description  of  ethos  and  was  written  after  I  had  dis- 
criminated the  two  types  of  schismogenesis. 


278  Epilogue  1936 

I  left  ethology  and  began  to  think  in  terms  of  developmental 
psychology. 

The  view  of  schismogenesis  provided  by  the  sketch  of 
latmul  eidos  v^^as  equally  limited  in  its  scope  and  I  think  it 
unlikely  that  I  should  ever,  from  this  aspect  alone,  have 
described  "  direct "  and  "  diagonal "  dualism  in  terms  of  com- 
plementary and  symmetrical  patterns  of  relationship.  Here 
the  difficulty  in  the  way  of  a  clear  phrasing  of  the  matter  was 
due  not  merely  to  the  lack  of  diachronic  perspective  but  also 
to  two  other  circumstances :  first,  I  had  already  attached  labels 
to  the  two  forms  of  dualism  and  it  would  have  been  difficult 
to  get  away  from  the  implications  of  these  labels ;  and  secondly, 
I  had  realised  that  it  was  immaterial  to  the  dualism  whether 
the  units  of  a  given  pair  were  identified  or  discriminated  and 
this  (illogically)  would  have  prevented  me  from  inspecting 
the  patterning  of  the  relationships. 

Thus  each  of  our  separate  methods  involves  distortions  and 
gives  but  a  partial  view  of  the  phenomena,  but  I  do  not  think 
that  for  this  reason  the  methods  should  be  discarded.  Against 
the  incompleteness  of  each  isolated  aspect  of  the  phenomena, 
we  may  set  the  increased  simplicity  which  such  isolation 
provides  and  the  advantages  of  being  able  to  state 
our  problems  and  generalisations  in  such  clear  terms  that 
we  can  at  once  look  for  the  relevant  facts.  Moreover,  it  may 
be  remembered  that  the  architect  who  draws  first  a  plan  and 
after  that  an  elevation  of  a  building  is  not  thereby  prevented 
from  finally  drawing  a  picture  of  the  building  in  perspective. 
After  we  have  looked  at  our  sketches  of  isolated  aspects  of 
the  culture,  it  will  still  be  possible  to  go  on  to  a  synthesis 
of  these  aspects,  a  synthesis  which  will  involve  its  own  types 
of  distortion  but  which  will  be  the  more  complete  because 
in  collecting  our  facts  we  shall  have  been  guided  by  several 
diff"erent  types  of  relevance. 

Finally,  it  remains  for  me  to  state  what  value  I  attach  to 
the  facts,  the  theories,  and  the  methods,  which  I  have  put 
forward.  It  is  clear  that  I  have  contributed  but  little  to  our 
store  of  anthropological  facts  and  that  the  information  about 
latmul  culture  which  I  have  used  in  the  various  chapters  does 
no  more  than  illustrate  my  methods.    Even  for  purposes  of 


Epilogue  1936  279 

illustration  my  supply  of  facts  is  meagre,  and  I  certainly 
cannot  claim  that  my  facts  have  demonstrated  the  truth  of 
any  theory. 

This  would  be  a  serious  confession  of  weakness,  and  indeed 
the  book  would  have  no  value,  were  it  not  for  another  short- 
coming which  in  a  sense  cancels  out  the  first.  To  have  put 
forward,  unsupported  by  a  solid  backing  of  fact,  theories  and 
hypotheses  which  were  new,  would  have  been  criminal,  but 
it  so  happens  that  none  of  my  theories  is  in  any  sense  new 
or  strange.  They  are  all  to  some  extent  platitudes,  which 
novelists,  philosophers,  religious  leaders,  lawyers,  the  man  in 
the  street,  and  even  anthropologists,  have  reiterated  in  various 
forms  probably  since  language  was  invented.  Structure, 
ethos,  and  the  rest  are  not  new  ideas  or  new  theories:  they 
are  only  new  labels  for  old  ways  of  thought,  and  indeed  only 
two  of  my  abstractions — eidos  and  schismogenesis — can  claim 
the  distinction  of  receiving  even  new  labels. 

I  am  aware  that  there  are  dangers  inherent  in  the  use  of 
labels,  and  that  these  little  pieces  of  paper  are  only  too  liable 
to  hide  the  things  to  which  they  are  attached.  But  though 
labels  must  always  be  used  with  caution,  they  are  useful 
things,  and  the  whole  of  science  depends  on  them.  In  this 
case,  their  use  has  been  to  help  me  to  disentangle  old  ideas 
and  to  enable  me  to  think  in  terms  of  one  aspect  of  culture 
at  a  time  instead  of  jumbling  all  the  aspects  together.  Our 
unscientific  knowledge  of  the  diverse  facets  of  human  nature 
is  prodigious,  and  only  when  this  knowledge  has  been  set  in 
a  scientific  framework  shall  we  be  able  to  hope  for  new  ideas 
and  theories. 


CHAPTER   XVII 

Epilogue  1958 

THERE  is  a  well-known  story  about  the  philosopher 
Whitehead.  His  former  pupil  and  famous  collaborator, 
Bertrand  Russell,  came  to  visit  Harvard  and  lectured  in 
the  large  auditorium  on  quantum  theory,  always  a  difficult  sub- 
ject, and  at  that  time  a  comparatively  novel  theory.  Russell 
labored  to  make  the  matter  intelligible  to  the  distinguished 
audience,  many  of  whom  were  unversed  in  the  ideas  of  mathe- 
matical physics.  When  he  sat  down,  Whitehead  rose  as  chair- 
man to  thank  the  speaker.  He  congratulated  Russell  on  his 
brilliant  exposition  "and  especially  on  leaving  .  .  .  unob- 
scured  .  .  .  the  vast  darkness  of  the  subject." 

All  science  is  an  attempt  to  cover  with  explanatory  de- 
vices—and thereby  to  obscure— the  vast  darkness  of  the  sub- 
ject. It  is  a  game  in  which  the  scientist  uses  his  explanatory 
principles  according  to  certain  rules  to  see  if  these  principles 
can  be  stretched  to  cover  the  vast  darkness.  But  the  rules  of  the 
stretching  are  rigorous,  and  the  purpose  of  the  whole  operation 
is  really  to  discover  what  parts  of  the  darkness  still  remain, 
uncovered  by  explanation. 

But  this  game  has  also  a  deeper,  more  philosophic  purpose: 
to  learn  something  about  the  very  nature  of  explanation,  to 
make  clear  some  part  of  that  most  obscure  matter— the  process 
of  knowing. 

In  the  twenty-one  years  that  have  elapsed  since  the  writing 
of  this  book,  epistemology— that  science  or  philosophy  which 
has  for  subject  matter  the  phenomena  which  we  call  knowl- 
edge and  explanation— has  undergone  an  almost  total  change. 
Preparing  the  book  for  repubhcation  in  1957  has  been  a  voyage 
of  discovery  backwards  into  a  period  when  the  newer  ways  of 
thought  were  only  dimly  foreshadowed. 

Naven  was  a  study  of  the  nature  pf  explanation.  The  book 
contains  of  course  details  about  latmul  life  and  culture,  but  it 
is  not  primarily  an  ethnographic  study,  a  retailing  of  data  for 


Epilogue  1958  28 1 

later  synthesis  by  other  scientists.  Rather,  it  is  an  attempt  at 
synthesis,  a  study  of  the  ways  in  which  data  can  be  fitted  to- 
gether, and  the  fitting  together  of  data  is  what  I  mean  by 
"explanation." 

The  book  is  clumsy  and  awkward,  in  parts  almost  unread- 
able. For  this  reason:  that,  when  I  wrote  it,  I  was  trying  not 
only  to  explain  by  fitting  data  together  but  also  to  use  this 
explanatory  process  as  an  example  within  which  the  princi- 
ples of  explanation  could  be  seen  and  studied. 

The  book  is  a  weaving  of  three  levels  of  abstraction.  At  the 
most  concrete  level  there  are  ethnographic  data.  More  ab- 
stract is  the  tentative  arranging  of  data  to  give  various  pictures 
of  the  culture,  and  still  more  abstract  is  the  self-conscious  dis- 
cussion of  the  procedures  by  which  the  pieces  of  the  jigsaw 
puzzle  are  put  together.  The  final  climax  of  the  book  is  the 
discovery,  described  in  the  epilogue— and  achieved  only  a  few 
days  before  the  book  went  to  press— of  what  looks  Hke  a  truism 
today:  that  ethos,  eidos,  sociology,  economics,  cultural  struc- 
ture, social  structure,  and  all  the  rest  of  these  words  refer  only 
to  scientists'  ways  of  putting  the  jigsaw  puzzle  together. 

These  theoretical  concepts  have  an  order  of  objective  real- 
ity. They  are  really  descriptions  of  processes  of  knowing, 
adopted  by  scientists,  but  to  suggest  that  "ethos"  or  "social 
structure"  has  more  reality  than  this  is  to  commit  Whitehead's 
fallacy  of  misplaced  concreteness.  The  trap  or  illusion— like 
so  many  others— disappears  when  correct  logical  typing  is 
achieved.  If  "ethos,"  "social  structure,"  "economics,"  etc.,  are 
words  in  that  language  which  describes  how  scientists  arrange 
data,  then  these  words  cannot  be  used  to  "explain"  phenomena, 
nor  can  there  be  any  "ethological"  or  "economic"  categories  of 
phenomena.  People  can  be  influenced,  of  course,  by  economic 
theories  or  by  economic  fallacies— or  by  hunger— but  they  can- 
not possibly  be  influenced  by  "economics."  "Economics"  is  a 
class  of  explanations,  not  itself  an  explanation  of  anything. 

Once  the  fallacy  has  been  detected,  the  way  is  open  for 
growth  of  an  entirely  new  science— which  has  in  fact  already 
become  basic  to  modern  thought.  This  new  science  has  as  yet 
no  satisfactory  name.  A  part  of  it  is  included  within  what  is 
now  called  communications  theory,  a  part  of  it  is  in  cyber- 


282  Epilogue  1958 

netics,  a  part  of  it  in  mathematical  logic.  The  whole  is  still 
unnamed  and  imperfectly  envisioned.  It  is  a  new  structuring 
of  the  balance  between  Nominahsm  and  Reahsm,  a  new  set 
of  conceptual  frames  and  problems,  replacing  the  premises  and 
problems  set  by  Plato  and  Aristotle. 

One  purpose,  then,  of  the  present  essay  is  to  relate  the  book 
to  these  new  ways  of  thought  which  were  dimly  foreshadowed 
in  it.  A  second  more  specific  purpose  is  to  relate  the  book  to 
current  thinking  in  the  field  of  psychiatry.  While  the  climate 
of  epistemological  thought  has  been  changing  and  evolving 
throughout  the  world,  the  thinking  of  the  author  has  under' 
gone  changes,  precipitated  especially  by  contact  with  some  of 
the  problems  of  psychiatry.  I  have  had  the  task  of  teaching 
cultural  anthropology  to  psychiatric  residents,  and  have  had  to 
face  such  problems  as  are  raised  by  the  comparison  between 
the  variety  of  cultures  and  those  hazily  defined  "clinical  en- 
tities," the  mental  diseases  which  have  their  roots  in  traumatic 
experience. 

This  narrower  purpose,  to  make  the  book  relevant  to  psy- 
chiatry, is  easier  to  achieve  than  the  wider  one  of  placing  it  in 
the  current  epistemological  scene.  And  therefore  I  will  at- 
tempt the  psychiatric  problem  first,  with  this  reminder  to  the 
reader— that  the  problems  of  psychiatry  are  after  all  shot 
through  with  epistemological  difficulties. 

Naven  was  written  almost  without  benefit  of  Freud.  One 
or  two  reviewers  even  complained  about  this,  but  I  think  that 
the  circumstance  was  fortunate.  My  psychiatric  taste  and 
judgment  were  at  that  time  defective,  and  probably  a  greater 
contact  with  the  Freudian  ideas  would  only  have  led  me  to 
misuse  and  misunderstand  them.  I  would  have  indulged  in  an 
orgy  of  interpreting  symbols,  and  this  would  have  distracted 
me  from  the  more  important  problems  of  interpersonal  and 
intergroup  process.  As  it  was,  I  did  not  even  notice  that  the 
crocodile  jaw  which  is  the  gate  to  the  initiatory  enclosure  is 
called  in  latmul  tshuwi  iamfoa— literally,  "clitoris  gate."  This 
piece  of  data  would  really  only  confirm  what  is  already  implied 
when  the  male  initiators  are  identified  as  "mothers"  of  the 
novices,  but  still  the  temptation  to  analyze  the  symbolism  could 
have  interrupted  the  analysis  of  the  relationship. 


Epilogue  1958  283 

But  the  fascination  of  symbol  analysis  is  not  the  only  pitfall 
in  psychiatric  theory.  Perhaps  even  more  serious  are  the  dis- 
tractions of  psychological  typology.  One  of  the  great  errors  in 
anthropology  has  been  the  naive  attempt  to  use  psychiatric 
ideas  and  labels  to  explain  cultural  diflFerence,  and  certainly 
the  weakest  part  of  the  book  is  that  chapter  in  which  I  tried 
to  describe  ethological  contrast  in  terms  of  Kretschmer's 
typology. 

No  doubt  more  modern  approaches  to  the  problem  of  ty- 
pology, such  as  Sheldon's  work  on  somatotypes,  are  a  great 
improvement  upon  the  crudely  dualistic  system  of  Kretschmer. 
But  this  is  not  the  point  which  concerns  me.  If  Sheldon's 
typology  had  been  available  to  me  in  1935,  I  would  have  used 
it  in  preference  to  that  of  Kretschmer,  but  I  would  still  have 
been  wrong.  As  I  see  it  today,  these  typologies,  whether  in 
cultural  anthropology  or  in  psychiatry,  are  at  best  heuristic 
fallacies,  culs  de  sac,  whose  only  usefulness  is  to  demonstrate 
the  need  for  a  fresh  start.  Fortunately,  I  separated  my  dal- 
liance with  psychiatric  typology  into  a  single  chapter;  if  this 
were  not  so,  I  would  hardly  permit  the  republication  today. 

But  still  the  status  of  typology  is  undefined  and  crucial. 
Psychiatrists  still  hanker  after  classifications  of  mental  disease, 
biologists  still  hanker  for  genera  and  species;  and  physiologists 
still  hanker  after  a  classification  of  human  individuals  which 
shall  show  a  coincidence  between  classes  defined  by  behavioral 
criteria  and  classes  defined  by  anatomy.  Lastly,  be  it  con- 
fessed, I  myself  hanker  for  a  classification,  a  typology,  of  the 
processes  of  interaction  as  it  occurs  either  between  persons  or 
between  groups. 

This  is  a  region  in  which  problems  of  epistemology  become 
crucial  for  the  whole  biological  field,  including  within  that 
field  both  the  latmul  culture  and  psychiatric  diagnosis.  There 
is  an  area  of  comparable  uncertainty  in  the  whole  theory  of 
evolution:  Do  species  have  real  existence  or  are  they  only  a 
device  of  description?  How  are  we  to  resolve  the  old  contro- 
versies between  continuity  and  discontinuity?  Or  how  shall 
we  reconcile  the  contrast  which  recurs  again  and  again  in 
nature  between  continuity  of  change  and  discontinuity  of  the 
classes  which  result  from  change? 


284  Epilogue  1958 

It  seems  to  me,  today,  that  there  is  a  partial  answer  to  these 
problems  in  the  processes  of  schismogenesis  which  are  ana- 
lyzed in  this  book,  but  this  partial  answer  could  hardly  have 
beeen  extracted  from  that  analysis  when  the  book  was  written. 
These  further  steps  had  to  wait  upon  other  advances,  such  as 
the  expansion  of  learning  theory,  the  development  of  cyber- 
netics, the  apphcation  of  Russell's  Theory  of  Logical  Types  to 
communications  theory,  and  Ashby's  formal  analysis  of  those 
orders  of  event  which  must  lead  to  parametric  change  in  pre- 
viously steady-state  systems. 

A  discussion  of  the  relationship  between  schismogenesis 
and  these  more  modern  developments  of  theory  is  therefore  a 
first  step  toward  a  new  synthesis.  In  this  discussion  I  shall 
assume  that  formal  analogies  exist  between  the  problems  of 
change  in  all  fields  of  biological  science. 

The  process  of  schismogenesis,  as  described  in  the  book,  is 
an  example  of  progressive  or  directional  change.  And  a  first 
problem  in  evolution  is  that  of  direction.  The  conventional 
stochastic  view  of  mutation  assumes  that  change  will  be  ran- 
dom, and  that  direction  is  only  imposed  upon  evolutionary 
change  by  some  phenomenon  hke  natural  selection.  Whether 
such  a  description  is  sufiicient  to  explain  the  phenomenon  of 
orthogenesis— the  long  process  of  continuous  directional 
change  shown  by  the  fossil  record  in  ammonites,  sea  urchins, 
horses,  asses,  titanotheres,  etc.— is  very  doubtful.  An  alterna- 
tive or  supplementary  explanation  is  probably  necessary.  Of 
these  the  most  obvious  is  climatic  or  other  progressive  change 
in  the  environment,  and  this  type  of  explanation  may  be  appro- 
priate for  some  of  the  orthogenic  sequences.  More  interesting 
is  the  possibility  that  the  progressive  environmental  change 
might  occur  in  the  biological  environment  of  the  species  con- 
cerned, and  this  raises  questions  of  a  new  order.  Marine  organ- 
isms like  ammonites  or  sea  urchins  can  hardly  be  supposed  to 
have  much  efifect  upon  the  weather,  but  a  change  in  the  am- 
monites might  affect  their  biological  environment.  After  all, 
the  most  important  elements  in  the  environment  of  an  indi- 
vidual organism  are  ( a )  other  individuals  of  the  same  species 
and  (b)  plants  and  animals  of  other  species  with  which  the 
given  individual  is  in  intense  interactive  relationship.    The 


Epilogue  1958  285 

survival  value  of  a  given  characteristic  is  likely  to  depend  in 
part  upon  the  degree  to  which  this  characteristic  is  shared  by 
other  members  of  the  species;  and,  vis-a-vis  other  species,  there 
must  exist  relationships— e.g.,  between  predator  and  prey— 
which  are  comparable  to  those  evolving  interactive  systems  of 
attack  and  defense  so  grievously  familiar  in  armaments  races 
at  the  international  level. 

These  are  systems  which  begin  to  be  closely  comparable  to 
the  phenomena  of  schismogenesis  with  which  this  book  is  con- 
cerned. In  the  theory  of  schismogenesis,  however  (and  in 
armaments  races ) ,  an  additional  factor  to  account  for  the  di- 
rectedness  of  change  is  assumed.  The  direction  toward  more 
intense  rivalry  in  the  case  of  symmetrical  schismogenesis,  or 
toward  increasing  differentiation  of  role  in  complementary 
schismogenesis,  is  assumed  to  depend  upon  phenomena  of 
learning.  This  aspect  of  the  matter  is  not  discussed  in  the  book, 
but  the  whole  theory  rests  upon  certain  ideas  about  processes 
of  character  formation— ideas  which  are  also  latent  in  most  psy- 
chiatric theory.   These  ideas  may  be  briefly  summarized. 

The  order  of  learning  to  which  I  refer  is  that  which  Harlow 
has  called  "set-learning,"  and  which  I  myself  have  called  "deu- 
tero-learning."  I  assume  that  in  any  learning  experiment— 
e.g.,  of  the  Pavlovian  or  the  Instrumental  Reward  type— there 
occurs  not  only  that  learning  in  which  the  experimenter  is 
usually  interested,  namely,  the  increased  frequency  of  the  con- 
ditioned response  in  the  experimental  context,  but  also  a  more 
abstract  or  higher  order  of  learning,  in  which  the  experimental 
subject  improves  his  ability  to  deal  with  contexts  of  a  given 
type.  The  subject  comes  to  act  more  and  more  as  if  contexts 
of  this  type  were  expectable  in  his  universe.  For  example,  the 
deutero-learning  of  the  animal  subjected  to  a  sequence  of  Pav- 
lovian experiences  will  presumably  be  a  process  of  character 
formation  whereby  he  comes  to  Hve  as  if  in  a  universe  where 
premonitory  signs  of  later  reinforcements  (or  unconditioned 
stimuli )  can  be  detected  but  nothing  can  be  done  to  precipi- 
tate or  prevent  the  occurrence  of  the  reinforcements.  In  a 
word,  the  animal  will  acquire  a  species  of  "fatalism."  In  con- 
trast, we  may  expect  the  subject  of  repeated  Instrumental 
Reward  experiments  to  deutero-learn  a  character  structure 


286  Epilogue  1958 

that  will  enable  him  to  live  as  if  in  a  universe  in  which  he  can 
control  the  occurrence  of  reinforcements. 

Now,  all  those  psychiatric  theories  which  invoke  the  past 
experience  of  the  individual  as  an  explanatory  device  depend 
necessarily  upon  some  such  theory  of  high-order  learning,  or 
learning  to  learn.  When  the  patient  tells  the  therapist  that,  in 
her  childhood,  she  learned  to  operate  a  typewriter,  this  is  of 
no  particular  interest  to  him  unless  he  happens  to  be  a  voca- 
tional counselor  as  well  as  therapist.  But  when  she  starts  to 
tell  him  about  the  context  in  which  she  learned  this  skill,  how 
her  aunt  taught  her  and  rewarded  her  or  punished  her  or  with- 
held reward  and  punishment,  then  the  psychiatrist  begins  to  be 
interested;  because  what  the  patient  learned  from  formal 
characteristics  or  patterns  of  the  contexts  of  learning  is  the 
clue  to  her  present  habits,  her  "character,"  her  manner  of  inter- 
preting and  participating  in  the  interaction  between  herself 
and  others. 

This  type  of  theory  which  underHes  so  much  of  psychiatry 
is  also  fundamental  to  the  idea  of  schismogenesis.  It  is  as- 
sumed that  the  individual  in  a  symmetrical  relationship  with 
another  will  tend,  perhaps  unconsciously,  to  form  the  habit  of 
acting  as  if  he  expected  symmetry  in  further  encounters  with 
that  other,  and  perhaps,  even  more  widely,  in  future  encoun- 
ters with  all  other  individuals. 

The  ground  is  thus  laid  for  progressive  change.  As  a  given 
individual  learns  patterns  of  symmetrical  behavior  he  not  only 
comes  to  expect  this  type  of  behavior  in  others,  but  also  acts  in 
such  a  way  that  others  will  experience  those  contexts  in  which 
they  in  turn  will  learn  symmetrical  behavior.  We  have  here  a 
case  in  which  change  in  the  individual  aflFects  the  environment 
of  others  in  a  way  which  will  cause  a  similar  change  in  them. 
This  will  act  back  upon  the  initial  individual  to  produce  fur- 
ther change  in  him  in  the  same  direction. 

But  this  picture  of  schismogenesis  cannot  be  true  of  latmul 
society  as  I  observed  it.  Evidently,  what  has  been  achieved  is 
only  a  one-sided  picture  of  processes  which,  if  permitted, 
would  lead  either  in  the  direction  of  excessive  rivalry  between 
symmetrical  pairs  or  groups  of  individuals  or  in  the  direction  of 
excessive  diSerentiation  between  complementary  pairs.  At  a 


Epilogue  1958  287 

certain  point,  if  these  were  the  only  processes  involved,  the 
society  would  explode.  Of  this  diflBculty  I  was  already  aware 
when  I  wrote  the  book,  and  I  made  an  effort  to  account  for  the 
presumed  dynamic  equilibrium  of  the  system  by  pointing  out 
that  the  symmetrical  and  complementary  processes  are  in  some 
sense  opposites  of  each  other  (page  193)  so  that  the  culture 
containing  both  of  these  processes  might  conceivably  balance 
them  one  against  the  other.  This,  however,  was  at  best  an 
unsatisfactory  explanation,  since  it  assumed  that  two  variables 
will,  by  coincidence,  have  equal  and  opposite  values;  but  it  is 
obviously  improbable  that  the  two  processes  will  balance  each 
other  unless  some  functional  relationship  obtains  between 
them.  In  the  so-called  dynamic  equilibrium  of  chemical  re- 
actions, the  rate  of  change  in  one  direction  is  a  function  of  the 
concentration  of  the  products  of  the  inverse  change,  and  re- 
ciprocally. But  I  was  not  able  to  see  any  such  functional  de- 
pendence between  the  two  schismogenic  processes  and  had  to 
leave  the  matter  there  when  the  book  was  written. 

The  problem  became  totally  changed  with  the  growth  of 
cybernetic  theory.  It  was  my  privilege  to  be  a  member  of  the 
Macy  Conference  which  met  periodically  in  the  years  follow- 
ing the  end  of  World  War  II.  In  our  earlier  meetings  the  word 
"cybernetics"  was  still  uncoined,  and  the  group  was  gathered 
to  consider  the  implications  for  biology  and  other  sciences  of 
what  we  then  called  "feed-back."  It  was  immediately  evident 
that  the  whole  problem  of  purpose  and  adaptation— the  teleo- 
logical  problem  in  the  widest  sense— had  to  be  reconsidered. 
The  problems  had  been  posed  by  the  Greek  philosophers,  and 
the  only  solution  they  had  been  able  to  offer  was  what  looked 
like  a  mystical  idea:  that  the  end  of  a  process  could  be  re- 
garded as  a  "purpose,"  and  that  this  end  or  purpose  could  be 
invoked  as  an  explanation  of  the  process  which  preceded  it. 
And  this  notion,  as  is  well  known,  was  closely  connected  with 
the  problem  of  the  real  (i.e.,  transcendent  rather  than  imma- 
nent) nature  of  forms  and  patterns.  The  formal  study  of  feed- 
back systems  immediately  changed  all  this.  Now,  we  had 
mechanical  models  of  causal  circuits  which  would  (if  the 
parameters  of  the  system  were  appropriate )  seek  equilibria  or 
steady  states.    But  Naven  had  been  written  with  a  rigorous 


288  Epilogue  1958 

taboo  on  teleological  explanation:  the  end  could  never  be  in- 
voked as  an  explanation  of  the  process. 

The  idea  of  negative  feedback  was  not  new;  it  had  been 
used  by  Clark  Maxwell  in  his  analysis  of  the  steam  engine  with 
a  governor,  and  by  biologists  such  as  Claude  Bernard  and 
Cannon  in  the  explanation  of  physiological  homeostasis.  But 
the  power  of  the  idea  was  unrecognized.  What  happened  at 
the  Macy  meetings  was  an  exploration  of  the  enormous  scope 
of  these  ideas  in  the  explanation  of  biological  and  social  phe- 
nomena. 

The  ideas  themselves  are  extremely  simple.  All  that  is  re- 
quired is  that  we  ask  not  about  the  characteristics  of  lineal 
chains  of  cause  and  effect  but  about  the  characteristics  of  sys- 
tems in  which  the  chains  of  cause  and  effect  are  circular  or 
more  complex  than  circular.  If,  for  example,  we  consider  a 
circular  system  containing  elements  A,  B,  C,  and  D— so  related 
that  an  activity  of  A  affects  an  activity  of  B,  B  affects  C,  C  af- 
fects D,  and  D  has  an  effect  back  upon  A— we  find  that  such  a 
system  has  properties  totally  different  from  anything  which 
can  occur  in  lineal  chains. 

Such  circular  causal  systems  must  in  the  nature  of  the  case 
either  seek  a  steady  state  or  undergo  progressive  exponential 
change;  and  this  change  will  be  limited  either  by  the  energy 
resources  of  the  system,  or  by  some  external  restraint,  or  by  a 
breakdown  of  the  system  as  such. 

The  steam  engine  with  the  governor  illustrates  the  type  of 
circuit  which  may  seek  a  steady  state.  Here  the  circuit  is  so 
constructed  that  the  faster  the  piston  moves  the  faster  the 
governor  spins;  and  the  faster  the  governor  spins  the  wider 
the  divergence  of  its  weighted  arms;  and  the  wider  the  diver- 
gences of  these  arms  the  less  the  power  supply.  But  this  in 
turn  affects  the  activity  of  the  piston.  The  self -corrective  char- 
acteristic of  the  circuit  as  a  whole  depends  upon  there  being 
within  the  circuit  at  least  one  link  such  that  the  more  there  is 
of  something,  the  less  there  will  be  of  something  else.  In  such 
cases  the  system  may  be  self-corrective,  either  seeking  a  steady 
rate  of  operation  or  oscillating  about  such  a  steady  rate. 

In  contrast,  a  steam  engine  with  a  governor  so  constructed 
that  a  wider  divergence  of  the  arms  of  the  governor  will  in- 


Epilogue  1958  289 

crease  the  supply  of  steam  to  the  cyhnder  aflFords  an  instance 
of  what  the  engineers  would  call  "runaway."  The  feedback  is 
"positive,"  and  the  system  will  operate  faster  and  faster,  in- 
creasing its  speed  exponentially  to  the  limit  of  the  available 
supply  of  steam  or  to  the  point  at  which  the  flywheel  or  some 
other  part  must  break. 

For  the  present  purposes  it  is  not  necessary  to  go  into  the 
mathematics  of  such  systems  except  to  notice  that  the  charac- 
teristics of  any  such  system  will  depend  upon  timing.  Will  the 
corrective  event  or  message  reach  the  point  at  which  it  is  effec- 
tive at  an  appropriate  moment  and  will  the  effect  be  sufficient? 
Or  will  the  corrective  action  be  excessive?  Or  too  httle?  Or 
too  late? 

Substituting  the  notion  of  seff-correction  for  the  idea  of 
purpose  or  adaptation  defined  a  new  approach  to  the  problems 
of  the  latmul  culture.  Schismogenesis  appeared  to  promote 
progressive  change,  and  the  problem  was  why  this  progressive 
change  did  not  lead  to  a  destruction  of  the  culture  as  such. 
With  self-corrective  causal  circuits  as  a  conceptual  model,  it 
was  now  natural  to  ask  whether  there  might  exist,  in  this  cul- 
ture, functional  connections  such  that  appropriate  controlling 
factors  would  be  brought  into  play  by  increased  schismo- 
genic  tension.  It  was  not  good  enough  to  say  that  symmetrical 
schismogenesis  happened  by  coincidence  to  balance  the 
complementary.  It  was  now  necessary  to  ask,  is  there  any 
communicational  pathway  such  that  an  increase  in  sym- 
metrical schismogenesis  will  bring  about  an  increase  in  the 
corrective  complementary  phenomena?  Could  the  system  be 
circular  and  self-corrective? 

The  answer  was  immediately  evident  ( page  8 ) .  The  naven 
ceremonial,  which  is  an  exaggerated  caricature  of  a  comple- 
mentary sexual  relationship  between  wau  and  laua,  is  in  fact 
set  off  by  overweening  symmetrical  behavior.  When  laua 
boasts  in  the  presence  of  wau,  the  latter  has  recourse  to  naven 
behavior.  Perhaps  in  the  initial  description  of  the  contexts  for 
naven  it  would  have  been  better  to  describe  this  as  the  primary 
context,  and  to  see  laua's  achievements  in  headhunting,  fishing, 
etc.,  as  particular  examples  of  an  achieved  ambition  or  vertical 
mobility  in  laua  which  place  him  in  some  sort  of  symmetrical 


290  Epilogue  1958 

relationship  with  wau.  But  the  latmul  do  not  think  of  the  mat- 
ter this  way.  If  you  ask  a  latmul  about  the  contexts  for  naven, 
he  will  first  enumerate  laua's  achievements  and  only  as  an 
afterthought  mention  the  less  formal  ( but  perhaps  more  pro- 
foundly significant )  contexts  in  which  wau  uses  naven  to  con- 
trol that  breach  of  good  manners  of  which  laua  is  guilty  when 
he  presumes  to  be  in  a  symmetrical  relationship  with  wau. 
Indeed,  it  was  only  on  a  later  visit  to  latmul  that  I  discovered 
that  when  laua  is  a  baby  and  is  being  held  in  waus  lap,  if  the 
baby  urinates,  the  wau  will  threaten  naven. 

It  is  also  interesting  that  this  link  between  symmetrical  and 
complementary  behavior  is  doubly  inverted.  The  laua  makes 
the  symmetrical  gesture  and  wau  responds  not  by  overbearing 
complementary  dominance  but  by  the  reverse  of  this— exag- 
gerated submission.  Or  should  we  say  the  reverse  of  this  re- 
verse? Waus  behavior  is  a  caricature  of  submission? 

The  sociological  functions  of  this  self-corrective  circuit 
cannot  be  so  easily  demonstrated.  The  questions  at  issue  are 
whether  excessive  symmetrical  rivalry  between  clans  will  in 
fact  increase  the  frequency  with  which  lauas  act  symmetrically 
yls-a^vis  their  waus,  and  whether  the  resulting  increase  in  fre- 
quency of  naven  will  tend  to  stabilize  the  society.  This  could 
only  be  demonstrated  by  statistical  study  and  appropriate 
measurement,  which  would  be  extremely  diflBcult.  There  is, 
however,  a  good  case  for  expecting  such  effects  inasmuch  as 
the  wau  is  usually  of  a  different  clan  from  laua.  In  any  instance 
of  intense  symmetrical  rivalry  between  two  clans  we  may  ex- 
pect an  increased  probability  of  symmetrical  insult  between 
members,  and  when  the  members  of  the  pair  happen  to  be 
related  as  laua  and  wau,  we  must  expect  a  triggering  of  the 
complementary  rituals  which  will  function  toward  mending 
the  threatened  split  in  society. 

But  if  there  exists  a  functional  relationship  such  that  excess 
of  symmetrical  rivahy  will  trigger  complementary  rituals,  then 
we  should  expect  to  find  also  the  converse  phenomenon.  In- 
deed, it  is  not  clear  that  the  society  could  maintain  its  steady 
state  without  an  excess  of  complementary  schismogenesis 
touching  off  some  degree  of  symmetrical  rivalry. 

This  too  can  be  demonstrated  with  ethnographic  data: 


Epilogue  1958  29 1 

( 1 )  In  the  village  of  Tambunum,  when  two  little  boys  ex- 
hibit what  looks  to  their  age  mates  like  homosexual  behavior, 
the  others  put  sticks  in  their  hands  and  make  the  two  stand  up 
against  each  other  and  "fight."  Indeed,  any  suggestion  of  pas- 
sive male  homosexuality  is  exceedingly  insulting  in  latmul 
culture  and  leads  to  symmetrical  brawling. 

(2)  As  discussed  in  the  book,  while  the  wau's  transvesti- 
cism  is  a  caricature  of  the  female  role,  the  transvesticism  of 
father's  sister  and  elder  brother's  wife  is  a  proud  exhibition 
of  masculinity.  It  looks  as  though  these  women  are  stating  a 
symmetrical  rivalry  vis-a-vis  the  men,  compensating  for  their 
normally  complementary  role.  It  is  perhaps  significant  that 
they  do  this  at  a  time  when  a  man,  the  wau,  is  stating  his  com- 
plementarity vis-a-vis  laua. 

( 3 )  The  extreme  complementarity  of  relationship  between 
initiators  and  novices  is  always  counterbalanced  by  extreme 
rivalry  between  initiatory  groups.  Here  again  complementary 
behavior  in  some  way  sets  the  scene  for  symmetrical  rivalry. 

Again  we  may  ask  the  sociological  question,  whether  these 
changes  from  complementarity  to  symmetry  can  be  regarded 
as  efiFective  in  the  prevention  of  social  disintegration;  and, 
again,  to  investigate  this  question  with  the  available  examples 
is  diflScult.  There  is,  however,  another  aspect  of  the  matter 
which  justifies  us  in  believing  that  this  oscillation  between  the 
symmetrical  and  the  complementary  is  likely  to  be  of  deep 
importance  to  the  social  structure.  What  has  been  demon- 
strated from  the  data  is  that  latmul  individuals  recurrently 
experience  and  participate  in  such  shifts.  From  this  we  may 
reasonably  expect  that  these  individuals  learn,  besides  the 
symmetrical  and  the  complementary  patterns,  to  expect  and 
exhibit  certain  sequential  relations  between  the  symmetrical 
and  the  complementary.  Not  only  must  we  think  of  a  social 
network  changing  from  moment  to  moment  and  impinging 
upon  the  individuals,  so  that  processes  tending  toward  dis- 
integration will  be  corrected  by  activation  of  other  processes 
tending  in  an  opposite  direction,  but  also  we  have  to  remem- 
ber that  the  component  individuals  of  that  network  are  them- 
selves being  trained  to  introduce  this  type  of  corrective  change 
in  their  dealings  with  each  other.  In  one  case  we  are  equating 


2Q2  Epilogue  1958 

the  individuals  with  the  A,  B,  C,  and  D  of  a  cybernetic  dia- 
gram; and  in  the  other,  noting  that  A,  B,  C,  etc.,  are  themselves 
so  structured  that  the  input-output  characteristics  of  each  will 
show  appropriate  self-corrective  characteristics. 

It  is  this  fact— that  the  patterns  of  society  as  a  major  entity 
can  by  learning  be  introjected  or  conceptualized  by  the  partici- 
pant individuals— that  makes  anthropology  and  indeed  the 
whole  of  behavioral  science  pecuharly  difficult.  The  scientist 
is  not  the  only  human  being  in  the  picture.  His  subjects  also 
are  capable  of  all  sorts  of  learning  and  conceptualization  and 
even,  like  the  scientist,  they  are  capable  of  errors  of  conceptual- 
ization. This  circumstance,  however,  leads  us  into  a  further  set 
of  questions  posed  by  communications  theory,  namely,  those 
questions  which  concern  the  orders  of  event  which  will  trigger 
corrective  action,  and  the  order  of  that  action  ( considered  as  a 
message)  when  it  occurs. 

Here  I  use  the  word  "order"  in  a  technical  sense  closely  re- 
sembling the  sense  in  which  the  word  "type"  is  used  in  Rus- 
sell's Theory  of  Logical  Types.  This  may  be  illustrated  by  the 
following  example.  A  house  with  a  thermostatically  controlled 
heating  system  is  a  simple  self-corrective  circuit  of  the  sort 
discussed  above.  A  thermometer  appropriately  placed  in  the 
house  is  linked  into  the  system  to  control  a  switch  in  such  a 
way  that  when  the  temperature  goes  above  a  certain  critical 
level  the  furnace  is  switched  oflF.  Similarly,  when  the  tempera- 
ture falls  below  a  certain  level  the  furnace  is  switched  on.  But 
the  system  is  also  governed  by  another  circumstance,  namely, 
the  setting  of  the  critical  temperatures.  By  changing  the  posi- 
tion of  a  dial  the  owner  of  the  house  can  alter  the  character- 
istics of  the  system  as  a  whole  by  changing  the  critical  tem- 
peratures at  which  the  furnace  will  be  turned  on  and  shut  off. 
Following  Ashby,  I  will  reserve  the  word  "variables"  for  those 
measurable  circumstances  which  change  from  moment  to  mo- 
ment as  the  house  oscillates  around  some  steady  temperature, 
and  shall  reserve  the  word  "parameters"  for  those  character- 
istics of  the  system  which  are  changed  for  example  when  the 
householder  intervenes  and  changes  the  setting  of  the  thermo- 
stat. I  shall  speak  of  the  latter  change  as  of  higher  order  than 
changes  in  the  variables. 


Epilogue  1958  293 

The  word  "order"  is  in  fact  here  used  in  a  sense  comparable 
to  that  in  which  it  was  used  earher  in  this  essay  to  define  orders 
of  learning.  We  deal  as  before  with  meta  relationships  be- 
tween messages.  Any  two  orders  of  learning  are  related  so  that 
the  learning  of  one  order  is  a  learning  about  the  other,  and 
similarly  in  the  case  of  the  house  thermostat  the  message  which 
the  householder  puts  into  the  system  by  changing  the  setting 
is  about  how  the  system  shall  respond  to  messages  of  lower 
order  emanating  from  the  thermometer.  We  are  here  at  a  point 
where  both  learning  theory  and  the  theory  of  cybernetic  sys- 
tems come  within  the  realm  of  Russell's  Theory  of  Types. 

Russell's  central  notion  is  the  truism  that  a  class  cannot  be 
a  member  of  itself.  The  class  of  elephants  has  not  got  a  trunk 
and  is  not  an  elephant.  This  truism  must  evidently  apply  with 
equal  force  when  the  members  of  the  class  are  not  things  but 
names  or  signals.  The  class  of  commands  is  not  itself  a  com- 
mand and  cannot  tell  you  what  to  do. 

Corresponding  to  this  hierarchy  of  names,  classes,  and 
classes  of  classes,  there  is  also  a  hierarchy  of  propositions  and 
messages,  and  within  this  latter  hierarchy  the  Russellian  dis- 
continuity between  types  must  also  obtain.  We  speak  of  mes- 
sages, of  meta-messages,  and  meta-meta-messages;  and  what 
I  have  called  deutero-leaming  I  might  appropriately  have 
called  meta-learning. 

But  the  matter  becomes  more  diflBcult  because,  for  ex- 
ample, while  the  class  of  commands  is  not  itself  a  command, 
it  is  possible  and  even  usual  to  give  commands  in  a  meta-lan- 
guage.  If  "Shut  the  door"  is  a  command,  then  "Listen  to  my 
orders"  is  a  meta-command.  The  military  phrase,  "That  is  an 
order,"  is  an  attempt  to  enforce  a  given  command  by  appeal  to 
a  premise  of  higher  logical  type. 

Russell's  rule  would  indicate  that  as  we  should  not  classify 
the  class  of  elephants  among  its  members,  so  also  we  should 
not  classify  "Listen  to  my  orders"  among  such  commands  as 
"Shut  the  door."  But,  being  human,  we  shall  continue  to  do  so 
and  shall  inevitably  be  liable  to  certain  sorts  of  confusion,  as 
Russell  predicts. 

Returning  to  the  theme  which  I  am  trying  to  elucidate— the 
general  problem  of  the  continuity  of  process  and  the  discon- 


294  Epilogue  1958 

tinuity  of  the  products  of  process— I  will  now  consider  how  we 
might  classify  the  answers  to  this  general  problem.  Of  neces- 
sity, the  answers  will  be  in  the  most  general  terms,  but  still  it 
is  of  some  value  to  present  an  ordering  of  thoughts  about 
change  as  it  must,  a  priori,  occur  in  all  systems  or  entities 
which  either  learn  or  evolve/ 

First,  it  is  necessary  again  to  stress  the  distinction  between 
change  in  the  variables  (which  is,  by  definition,  within  the 
terms  of  the  given  system)  and  change  in  the  parameters,  i.e., 
change  in  the  very  terms  which  define  the  system— remember- 
ing always  that  it  is  the  observer  who  does  the  defining.  It  is 
the  observer  who  creates  messages  (i.e.,  science)  about  the 
system  which  he  is  studying,  and  it  is  these  messages  that  are 
of  necessity  in  some  language  or  other  and  must  therefore  have 
order:  they  must  be  of  some  or  other  Logical  Type  or  of  some 
combination  of  Types. 

The  scientist's  task  is  only  to  be  a  good  scientist,  that  is,  to 
create  his  description  of  the  system  out  of  messages  of  such 
logical  typology  ( or  so  interrelated  in  their  typology )  as  may 
be  appropriate  to  the  particular  system.  Whether  Russell's 
Types  "exist"  in  the  systems  which  the  scientist  studies  is  a 
philosophic  question  beyond  the  scientist's  scope— perhaps 
even  an  unreal  question.  For  the  scientist,  it  suffices  to  note 
that  logical  typing  is  an  inevitable  ingredient  in  the  relation- 
ship between  any  describer  and  any  system  to  be  described. 

What  I  am  proposing  is  that  the  scientist  should  accept  and 
use  this  phenomenon,  which  is,  in  any  case,  inevitable.  His 
science— the  aggregate  of  his  messages  about  the  system  which 
he  is  describing— will  be  so  constructed  that  it  could  be 
mapped  in  some  more  or  less  complex  diagram  of  logical  types. 
As  I  imagine  it,  each  message  of  the  description  would  have  its 

1  This  is  not  the  place  to  discuss  the  controversies  which  have  raged 
over  the  relation  between  learning  and  evolutionary  process.  Suffice  it 
that  two  contrasting  schools  of  thought  are  in  agreement  on  a  funda- 
mental analogy  between  the  two  genera  of  process.  On  the  one  hand, 
there  are  those  who,  following  Samuel  Butler,  argue  that  evolutionary 
change  is  a  sort  of  learning;  on  the  other  hand,  there  are  those  who  argue 
that  learning  is  a  sort  of  evolutionary  change.  Notable  among  the  latter 
are  Ashby  and  Mosteller,  whose  models  of  learning  involve  stochastic 
concepts  closely  comparable  to  the  concepts  of  natural  selection  and  ran- 
dom mutation. 


Epilogue  1958  295 

location  on  this  map,  and  the  topological  relationship  between 
various  locations  would  represent  the  typological  relationship 
between  messages.  It  is  of  the  nature  of  all  communication, 
as  we  know  it,  that  some  such  mapping  be  possible. 

But  in  describing  a  given  system,  the  scientist  makes  many 
choices.  He  chooses  his  words,  and  he  decides  which  parts  of 
the  system  he  will  describe  first;  he  even  decides  into  what 
parts  he  will  divide  the  system  in  order  to  describe  it.  These 
decisions  will  affect  the  description  as  a  whole  in  the  sense  that 
they  will  affect  the  map  upon  which  the  typological  relations 
between  the  elementary  messages  of  the  description  are  repre- 
sented. Two  equally  sufficient  descriptions  of  the  same  system 
could  conceivably  be  represented  by  utterly  different  map- 
pings. In  such  a  case,  is  there  any  criterion  by  which  it  would 
be  possible  for  the  scientist  to  choose  one  description  and  dis- 
card the  other? 

Evidently  an  answer  to  this  question  would  become  avail- 
able if  scientists  would  use,  as  well  as  accept,  the  phenomena 
of  logical  typing.  They  are  already  scrupulous  about  the  pre- 
cise coding  of  their  messages  and  insist  upon  singularity  of 
referent  for  every  symbol  used.  Ambiguity  at  this  simpler  level 
is  abhorred  and  is  avoided  by  rigorous  rules  for  the  translation 
of  observation  into  description.  But  this  rigor  of  coding  could 
also  be  useful  on  a  more  abstract  level.  The  typological  rela- 
tions between  the  messages  of  a  description  could  also  be 
used,  subject  to  rules  of  coding,  to  represent  relations  within 
the  system  to  be  described. 

After  all,  any  modification  of  the  signal  or  change  in  rela- 
tionship between  modifications  of  the  signal  can  be  made  to 
carry  a  message;  and  by  the  same  token  any  change  in  relation- 
ship between  messages  can  itself  carry  a  message.  There  is 
then  no  inherent  reason  why  the  various  species  of  meta  rela- 
tionship between  the  messages  of  our  description  should  not 
be  used  as  symbols  whose  referents  would  be  relationships 
within  the  system  to  be  described. 

Indeed,  something  like  this  technique  of  description  is  al- 
ready followed  in  certain  fields,  notably  in  the  equations  of 
motion.  Equations  of  first  order  (in  x)  denote  uniform  veloc- 
ity; equations  of  second  order  ( in  x^ )  imply  acceleration;  equa- 


2q6  Epilogue  1958 

tions  of  third  order  ( in  x^ )  imply  a  change  in  acceleration;  and 
so  on.  There  is  moreover  an  analogy  between  this  hierarchy 
of  equations  and  the  hierarchy  of  Logical  Types:  a  statement 
of  acceleration  is  vieta  to  a  statement  of  velocity.  The  f amihar 
Rule  of  Dimensions  is  to  physical  quantities  what  the  Theory 
of  Logical  Types  is  to  classes  and  propositions. 

I  am  suggesting  that  some  technique  of  this  sort  might  be 
used  in  describing  change  in  those  systems  which  either  learn 
or  evolve,  and  further  that  if  such  a  technique  were  adopted, 
there  would  then  be  a  natural  basis  upon  which  to  classify 
answers  to  the  problems  of  change  in  these  systems:  the  an- 
swers would  fall  into  classes  according  to  the  typology  of  the 
messages  which  they  contain.  And  this  classification  of  answers 
should  coincide  both  with  a  classification  of  systems  according 
to  their  typological  complexity  and  with  a  classification  of 
changes  according  to  their  orders. 

In  illustration  of  this,  it  is  now  possible  to  go  back  over  the 
whole  body  of  description  and  argument  contained  in  this 
book  and  to  dissect  it  on  a  generaHzed  typological  scale  or 
map. 

The  book  starts  with  two  descriptions  of  latmul  culture, 
in  each  of  which  relatively  concrete  observations  of  behavior 
are  used  to  validate  generaHzations.  The  "structural"  descrip- 
tion leads  to  and  validates  eidological  generalization,  and  a 
corpus  of  ethological  generalization  is  validated  by  observa- 
tions of  expression  of  affect. 

In  the  1936  Epilogue,  it  is  demonstrated  that  ethos  and 
eidos  are  only  alternative  ways  of  arranging  data  or  alternative 
"aspects"  of  the  data.  This,  I  believe,  is  another  way  of  saying 
that  these  generalizations  are  of  the  same  order  or  Russellian 
type.  I  needed,  for  obscure  reasons,  to  use  two  sorts  of  descrip- 
tion, but  the  presence  of  these  two  descriptions  does  not  de- 
note that  the  system  described  actually  has  a  complexity  of 
this  dual  nature. 

One  significant  duality  has  however  already  been  men- 
tioned in  this  brief  survey,  namely,  the  duality  between  obser- 
vations of  behavior  and  generahzation,  and  this  duality,  I 
believe,  here  reflects  a  special  complexity  in  the  system:  the 
dual  fact  of  learning  and  learning  to  learn.    A  step  in  Rus- 


Epilogue  1958  297 

sellian  typology  inherent  in  the  system  is  represented  by  a  cor- 
responding step  in  the  description. 

A  second  typological  contrast  in  the  description,  which  I 
believe  represents  a  real  contrast  in  the  system  described,  is 
that  between  ethos-eidos  on  the  one  hand  and  sociology  on 
the  other.  Here,  however,  the  matter  is  less  clear.  Insofar  as 
the  total  society  is  represented  in  native  thought  and  communi- 
cation, clearly  this  representation  is  of  higher  type  than  the 
representations  of  persons,  actions,  and  so  on.  It  would  follow 
that  a  segment  of  the  description  should  be  devoted  to  this 
entity,  and  that  the  delimiting  of  this  segment  from  the  rest  of 
the  description  would  represent  a  real  typological  contrast 
within  the  system  described.  But,  as  the  matter  is  presented 
in  the  book,  the  distinctions  are  not  perfectly  clear,  and  the 
idea  of  sociology  as  a  science  dealing  with  the  adaptation  and 
survival  of  societies  is  mixed  up  with  the  concept  "society"  as 
a  Gestalt  in  native  thought  and  communication. 

It  is  appropriate  next  to  ask  about  the  concept  "schismo- 
genesis."  Does  the  isolating  and  naming  of  this  phenomenon 
represent  an  extra  order  of  complexity  in  the  system? 

Here  the  answer  is  clearly  affirmative.  The  concept  "schis- 
mogenesis"  is  an  implicit  recognition  that  the  system  contains 
an  extra  order  of  complexity  due  to  the  combination  of  learn- 
ing vdth  the  interaction  of  persons.  The  schismogenic  unit  is 
a  two-person  subsystem.  This  subsystem  contains  the  poten- 
tialities of  a  cybernetic  circuit  which  might  go  into  progressive 
change;  it  cannot,  therefore,  be  conceptually  ignored  and  must 
be  described  in  a  language  of  higher  type  than  any  language 
used  to  describe  individual  behavior— the  latter  category  of 
phenomena  being  only  the  events  in  one  or  another  arc  of  the 
schismogenic  subsystem. 

It  is  necessary,  next,  to  note  that  the  original  description 
contained  a  major  error  in  its  typological  mapping.  The  de- 
scription is  presented  as  "synchronic,"^  which  is  more  modern 
terminology   may   be  translated   as   "excluding   irreversible 

1  There  is  also  a  second  sense  in  which  anthropologists  use  the  word 
"synchronic":  to  describe  a  study  of  cultiire  which  ignores  progressive 
change  by  considering  only  a  brief  (or  infinitesimal)  span  of  time.  In 
this  usage,  a  synchronic  description  differs  from  a  diachronic  ahnost  as 
differential  calculus  differs  from  integral  calculus. 


298  Epilogue  1958 

change."  The  basic  assumption  of  the  description  was  that 
the  system  described  was  in  a  steady  state,  such  that  all 
changes  within  it  could  be  regarded  as  changes  in  variables 
and  not  in  the  parameters.  In  self-justification,  I  may  claim 
that  I  stated  that  there  must  exist  factors  which  would  control 
the  runaways  of  schismogenesis— but  still  I  overlooked  what  is 
crucial  from  the  present  point  of  view:  that  the  system  must 
contain  still  larger  circuits  which  would  operate  correctively 
upon  the  schismogeneses.  In  omitting  to  make  this  deduc- 
tion, I  falsified  the  whole  logical  typology  of  the  description 
by  not  depicting  its  highest  level.  This  error  is  corrected  in  the 
earlier  part  of  the  present  Epilogue. 

It  is  thus  possible,  at  least  crudely,  to  examine  the  scientific 
description  of  a  system  and  to  relate  the  logical  typology  of  the 
description  to  the  circuit  structure  of  the  system  described. 
The  next  step  is  to  consider  descriptions  of  change  preparatory 
to  asking  how  a  classification  of  such  descriptions  may  be  re- 
lated to  problems  of  phenomenal  discontinuity. 

From  what  has  already  been  said,  it  is  clear  that  we  must 
expect  statements  about  change  to  be  always  in  a  language 
one  degree  more  abstract  than  the  language  which  would 
suffice  to  describe  the  steady  state.  As  statements  about  ac- 
celeration must  always  be  of  higher  logical  type  than  state- 
ments about  velocity,  so  also  statements  about  cultural  change 
must  be  of  higher  type  than  synchronic  statements  about  cul- 
ture. This  rule  will  apply  throughout  the  field  of  learning  and 
evolution.  The  language  for  the  description  of  character 
change  must  always  be  of  higher  type  than  the  description  of 
character;  the  language  to  describe  psychiatric  etiology  or  psy- 
chotherapy, both  of  which  impute  change,  must  always  be 
more  abstract  than  the  language  of  diagnosis.  And  so  on. 

But  this  is  only  another  way  of  saying  that  the  language 
appropriate  for  describing  change  in  a  given  system  is  that 
language  which  would  also  be  appropriate  for  describing  the 
top  typological  level  in  a  steady  state  system  having  one  more 
degree  of  complexity  in  its  circuits.  If  the  original  description 
of  latmul  culture,  as  put  forward  in  the  body  of  the  book,  had 
been  a  sufficient  and  correct  description  of  a  steady  state,  then 
the  language  of  the  additional  statements  about  the  larger  cir- 


Epilogue  1958  299 

cuits  would  have  been  precisely  that  language  appropriate  to 
describe  change  or  disturbance  in  that  steady  state. 

When  the  scientist  is  at  a  loss  to  find  an  appropriate  lan- 
guage for  the  description  of  change  in  some  system  which  he 
is  studying,  he  will  do  well  to  imagine  a  system  one  degree 
more  complex  and  to  borrow  from  the  more  complex  system 
a  language  appropriate  for  his  description  of  change  in  the 
simpler. 

Finally  it  becomes  possible  to  attempt  a  crude  listing  of 
types  of  change  and  to  relate  the  items  of  this  listing  to  the 
general  problem  from  which  I  started— that  of  the  contrast 
between  continuity  of  process  and  discontinuity  of  the  prod- 
ucts of  process. 

Take  for  a  starting  point  a  system  S  of  which  we  have  a 
description  with  given  complexity  C,  and  observe  that  the 
absolute  value  of  C  is  for  present  purposes  irrelevant.  We  are 
concerned  with  the  problem  of  change  and  not  at  all  with  abso- 
lute values. 

Consider  now  events  and  processes  within  S.  These  may 
be  classified  according  to  the  orders  of  statement  which  must 
be  made  in  the  description  of  S  in  order  to  represent  them.  The 
crucial  question  which  must  be  asked  about  any  event  or  proc- 
ess within  S  is:  Can  this  event  or  process  be  included  within  a 
description  of  S  as  a  steady  state  having  complexity  C?  If  it 
can  be  so  included,  all  is  well  and  we  are  not  dealing  with  any 
change  that  will  alter  the  parameters  of  the  system. 

The  more  interesting  case,  however,  is  that  in  which  events 
or  processes  are  noted  in  S  which  cannot  be  comprised  within 
the  steady-state  description  of  complexity  C.  We  then  face 
the  necessity  of  adding  some  sort  of  meta  description  to  be 
chosen  according  to  the  type  of  disturbance  which  we  observe. 

Three  types  of  such  disturbance  may  be  listed:  (a)  Pro- 
gressive change,  like  schismogenesis,  which  takes  place  in  the 
values  of  relatively  superficial  and  fast  changing  variables. 
This,  if  unchecked,  must  always  disrupt  the  parameters  of  the 
system.  ( b )  Progressive  change  which,  as  Ashby  has  pointed 
out,  must  always  occur  in  the  more  stable  variables  (or  pa- 
rameters? )  when  certain  of  the  more  superficial  variables  are 
controlled.  This  must  happen  whenever  limitation  is  imposed 


^00  Epilogue  1958 

upon  those  superficial  and  fast-changing  variables  which  pre- 
viously were  essential  Hnks  in  some  self -corrective  circuit.  An 
acrobat  must  always  lose  his  balance  if  he  is  unable  to  make 
changes  in  the  angle  between  his  body  and  his  balancing  pole. 

In  either  of  these  cases,  the  scientist  is  driven  to  add  to  his 
description  of  S  statements  of  higher  order  than  those  included 
within  the  previous  description  C. 

( c )  Lastly  there  is  the  case  of  "random"  events  occurring 
within  the  system  S.  These  become  especially  interesting 
when  a  degree  of  randomness  is  introduced  into  the  very  sig- 
nals upon  which  the  system  depends  for  its  self-corrective 
characteristics.  The  stochastic  theories  of  learning  and  the 
mutation-natural-selection  theory  of  evolution  both  invoke 
phenomena  of  this  kind  as  basis  for  description  or  explanation 
of  change.  The  stochastic  theories  of  learning  assume  random 
changes  of  some  kind  in  the  neurological  net,  while  the  muta- 
tion theory  assumes  random  changes  in  the  chromosomal  ag- 
gregate of  messages. 

In  terms  of  the  present  discussion  neither  of  these  theories 
is  satisfactory  because  both  leave  undefined  the  logical  typing 
of  the  word  "random."  We  must  expect,  a  priori,  that  the  ag- 
gregate of  messages  which  we  call  a  genotype  must  be  made 
up  of  individual  messages  of  very  various  typology,  carried 
either  by  individual  genes  or  by  constellations  of  genes.  It  is 
even  likely  that,  on  the  whole,  more  generalized  and  higher- 
type  messages  are  more  frequently  carried  by  constellations  of 
genes,  while  more  concrete  messages  are  in  general  carried  by 
individual  genes.  Of  this,  there  is  no  certain  knowledge,  but 
still  it  seems  improbable  that  small  "random"  disturbances  can 
alter  with  equal  frequency  messages  of  whatever  logical  type. 
W'e  must  ask,  then:  what  distribution  of  disturbance  among 
the  messages  of  various  types  do  the  proponents  of  these  theo- 
ries have  in  mind  when  they  use  the  word  "random"?  These, 
however,  are  questions  more  specific  than  the  broad  terms  of 
the  present  discussion  and  are  introduced  only  to  illustrate  the 
problems  which  are  posed  by  the  new  epistemology  which  is 
now  evolving. 

The  problem  of  discontinuity  now  falls  into  place  in  the 
sense  that  it  is  possible  to  classify  the  principal  types  of  process 


Epilogue  1958  ^01 

and  explanation  which  crystalHze  around  this  phenomenon. 
Consider  still  the  hypothetical  system,  S,  and  the  description 
of  this  system  whose  complexity  I  am  calling  C.  The  first  type 
of  discontinuity  is  that  relatively  trivial  case  in  which  the  state 
of  the  system  at  a  given  time  is  observed  to  contrast  with  its 
state  at  some  other  time,  but  where  the  differences  are  still 
such  as  can  be  subsumed  within  the  terms  of  the  given  descrip- 
tion. In  such  cases,  the  apparent  discontinuity  will  be  either 
an  artefact  resulting  from  the  spacing  of  our  observations  in 
time,  or  will  be  due  to  the  presence  of  on-off  phenomena  in  the 
communicational  mechanism  of  the  system  studied. 

A  less  trivial  case  is  proposed  in  considering  two  similar 
systems  Si  and  So,  both  of  them  undergoing  continuous  changes 
in  their  variables  such  that  the  two  systems  appear  to  be  di- 
verging or  becoming  more  and  more  different  one  from  the 
other.  Such  a  case  becomes  nontrivial  when  some  extra  factor 
is  involved  which  may  prevent  later  convergence.  But  any 
such  factor  will  of  course  have  to  be  represented  in  the  de- 
scription of  the  systems  by  messages  of  higher  logical  type. 

The  next  category  of  discontinuity  will  include  all  those 
cases  which  involve  parametric  contrast.  I  have  considered 
briefly  above  the  types  of  ongoing  process  which  must  lead  to 
parametric  disruption  and  have  noted  that  these  are  all  in- 
stances in  which  the  description  of  the  system  undergoing 
change  must  be  of  higher  logical  type  than  would  have  been 
the  case  in  absence  of  such  processes.  This  I  believe  to  be  true 
even  in  that  vast  majority  of  instances  where  the  parametric 
disturbance  leads  to  gross  simplification  of  parameters  after 
the  disruptive  change.  Most  commonly  such  disruptions  will- 
in  accordance  with  laws  of  probability— result  in  the  "death" 
of  the  system.  In  a  few  cases  some  simplified  version  of  S  may 
persist,  and,  in  still  fewer  cases,  the  parametric  disruption  will 
lead  to  the  creation  of  a  new  system,  typologically  more  com- 
plex than  the  original  S. 

It  is  this  rare  possibility  that  is  perhaps  most  fascinating  in 
the  whole  field  of  learning,  genetics,  and  evolution.  But,  while 
in  the  most  general  terms  it  is  possible  to  state  with  some  rigor 
what  sort  of  changes  are  here  envisaged  and  to  see  the  results 
of  such  progressive  discontinuous  change  in,  for  example,  the 


^02  Epilogue  1958 

telencephalization  of  the  mammalian  brain,  it  is  still  totally 
impossible  to  make  formal  statements  about  the  categories  of 
parametric  disturbance  which  will  bring  about  these  positive 
gains  in  complexity. 

Here  is  the  central  diflBculty  which  results  from  the  phe- 
nomenon of  logical  typing.  It  is  not,  in  the  nature  of  the  case, 
possible  to  predict  from  a  description  having  complexity  C 
what  the  system  would  look  like  if  it  had  complexity  C  +  1- 

This  formal  difficulty  must  in  the  end  always  limit  the 
scientific  understanding  of  change  and  must  at  the  same  time 
limit  the  possibilities  of  planned  change— whether  in  the  field 
of  genetics,  education,  psychotherapy,  or  social  planning. 

Certain  mysteries  are  for  formal  reasons  impenetrable,  and 
here  is  the  vast  darkness  of  the  subject. 


BIBLIOGRAPHY 

W.  R.  Ashby.   Design  for  a  Brain.    New  York:  Wiley,  1952. 

.  Introduction  to  Cybernetics.   New  York:  Wiley,  1956. 

G.  Bateson.  "Bali:  The  Value  System  of  a  Steady  State,"  in  M.  Fortes, 
ed.,  Social  Structure.   New  York:  Oxford  University  Press,  1949. 

.  "The  Message  'This  Is  Play,' "  in  Second  Conference  of  Group 

Processes.   New  York:  Josiah  Macy  Jr.  Foundation,  1956.   Pp.  145- 
242. 

G.  Bateson,  D.  D.  Jackson,  J.  Haley,  and  J.  H.  Weakland.  "Toward  a 
Theory  of  Schizophrenia,"  Behavioral  Science,  I  (1956),  251-64. 

R.  R.  Bush  and  F.  Mosteller.  Stochastic  Models  for  Learning.  New  York: 
Wiley,  1955. 

S.  Butler.   Life  and  Habit.   London:  Fifield,  1878. 

.  Luck  or  Cunning.   London:    Fifield,  1887. 

H.  von  Foerster,  ed.  Cybernetics.  5  vols.  New  York:  Josiah  Macy  Jr. 
Foundation,  1949-1953.  (Transactions  of  the  6di,  7th,  8th,  and  9th 
Conferences. ) 

C.  P.  Martin.  Psychology,  Evolution,  and  Sex.  Springfield,  111.:  Charles 
C.Thomas,  1956. 


303 

L.  F.  Richardson.  "Generalized  Foreign  Politics,"  British  Journal  of  Psy- 
chology Monograph  Supplements,  No.  23  (1939). 

J.  Ruesch  and  G.  Bateson.    Communication:  The  Social  Matrix  of  Psy- 
chiatry.  New  York:  W.  W.  Norton,  1951. 

B.  Russell.  Introduction  to  Wittgenstein,  Tractatus  Logico-Philosophicus. 

New  York:  Harcourt,  1956. 

C.  E.  Shannon  and  W.  Weaver.  The  Mathematical  Theory  of  Communi- 

cation.  Urbana:  University  of  Illinois  Press,  1949. 

C.  H.  Waddington.  The  Strategy  of  the  Genes.  London:  Allen  &  Unwin, 
1957. 

H.  Weyl.   Philosophy  of  Mathematics  and  Natural  Science.   Princeton: 
Princeton  University  Press,  1949. 

A.  N.  Whitehead  and  B.  Russell.    Principia  Mathematica.    Cambridge: 
Cambridge  University  Press,  1910. 

N.Wiener.  Cybernetics.  New  York:  Wiley,  1948. 


-*o 


-^ 


-'XD> 


I-'tJ 


O 

CH- 


Of 


I— 't5 


'X3 


-^ 


-'^ 


O 
N 

-Of 


o 


(-1 

-O 
O 
M 


-  o 


O 

N 

-O 

II    - 

o 

N 


9  ^ 


-'^ 


•+ 

^ 


—^> 


l_  o 

a 


4)    V 


O 

I— 0+- 


l-'^ 


-O 
O 
a 


«      — *o 


^ 


-^ 


(3    R 
C  .<3 


2  «< 


II 


S    O 


M       HH      H4 


I  3 

«  5 

4J  J)  fi 

o  o  i^i 

■u  '^  . 

PQ  '-'  t; 

w"''  3 

(3  ■w  t, 

5  o  c 

<3  aj  ' 

K  C  9 

•i  o  g 

a  2 


'^ 

C 

CO 

ea 

J, 

o 

« 

05 

(U 

0) 

"« 

3 

IS 

q-] 

'> 

<J 

^ 

fc 

C 

PQ 

•§  5 


3 


Xi     m  ■>*•  lo 


rj   5 
:   a   <3 


3  2 

&5   ^ 


O  •'5 

U    O 

o  en 
«  ^ 

5  S 


K    (3 


t^QO    O^    I— I  ' 


^- 


-»o 


o 

-04- 


'TD 


PQ 


— -  1) 

g2 

ndu. 

v    u 

§ai 

2  i2 

t}-  lOO 

o 

§. 

C    V 

S  - 

•g  j3 

C    3 

itnbo. 

CO    '-' 

•«  c 

nggw 

iai. 

mbuc 

II 

<3  <3      M    N    r^ 


INDEX   GLOSSARY   OF 
TECHNICAL  AND  NATIVE  WORDS 


Adaptive  function.  When  we  say  of  some  part  of  a  functional  system 
that  it  behaves  in  such  and  such  a  way  in  order  to  produce  such  and 
such  a  desirable  effect  in  the  system  as  a  whole,  we  are  attributing 
adaptive  function ;  and  we  are  verging  upon  teleological  fallacy.  But 
it  is  a  cold  plunge  from  this  to  realising  that  no  cell  or  organ  cares 
two  hoots  about  our  survival.  We  can  avoid  some  of  the  dangers  of 
teleology  by  acknowledging  that  undesirable  effects  occur  not  in- 
frequently. We  may  also  avoid  treating  the  teleological  theory  too 
seriously  by  saying  e.g.  that  the  father  plays  no  part  in  naven  because 
there  is  no  extra  integration  of  the  community  to  be  got  by  further 
stressing  of  the  links  between  father  and  son.    Cf.  pp.  27  and  211. 

Affective  and  cognitive.  These  terms  have  been  used  loosely  in  the  body 
of  the  book:  affective  to  mean  "pertaining  to  the  emotions"  and 
cognitive  to  mean  "pertaining  to  thought".  The  terms  are  re-inter- 
preted more  critically  on  p.  275. 

Affective  function.  The  effect  of  some  detail  of  cultural  behaviour  in 
satisfying  or  dissatisfying  the  emotional  needs  of  the  behaving  in- 
dividuals.   Cf.  p.  32. 

agwi,  islands  of  floating  grass,  which  may  attain  considerable  size  and 
solidity.  They  are  formed  on  lakes  but  float  down  to  the  Sepik  when 
broken  up.  Crocodiles  are  said  to  live  under  them.  Hence  wagan 
(who  are  crocodiles)  in  their  shamanic  jargon  refer  to  a  house  as  agwi. 

Alternation.   Cf.  p.  244  et  seq. 

angk-au,  literally  "potsherds",  hence  ancestral  spirits  sjmibolised  by 
the  old  potsherds  under  the  house.    Cf.  p.  45. 

awan,  the  suffix  of  names  given  by  mother's  brother  to  sister's  son.  The 
literal  meaning  of  this  suffix  is  either  "mask"  or  "old  man".    Cf. 

P-  43- 

bandi,  a  novice  or  initiate;  a  young  man;  a  member  of  initiatory  grade 
Ax  3  or  Ay  3 .  -bandi  is  a  suffix  on  names  given  by  wau  to  laua  after 
initiation.    Cf.  p.  43. 

b.s.  (both  speaking)  means  that  a  kinship  term  may  be  used  by  both 
sexes. 

Calculating  man.  The  individual  in  a  community  when  seen  from  a 
purely  economic  point  of  view.  This  phrase  need  make  no  assump- 
tions about  the  terms  in  which  the  man  "calculates".    Cf.  p.  27. 

Centripetal.  The  system  of  organisation  of  a  group  is  said  to  be  centri- 
petal if  it  depends  upon  a  single  central  authority  or  upon  some  form 
of  hierarchy.    Cf.  p.  99. 

Clan.  I  am  using  this  word  to  mean  a  unilateral  patrilineal  group.  Among 
the  latmul  there  are  between  fifty  and  a  hundred  clans  of  which  be- 
tween ten  and  twenty  are  represented  in  any  one  village.  In  many  cases 
the  clans  are  paired  and  trace  descent  from  a  pair  of  brothers,  one  clan 
being  called  the  "elder  brother"  of  the  other.  Larger  groups  of 
clans  also  occur.    See  Moiety. 

Cognitive.   Cf.  Affective. 


3o8  Glossary 

Complementary.  A  relationship  between  two  individuals  (or  between 
two  groups)  is  said  to  be  chiefly  complementary  if  most  of  the  be- 
haviour of  the  one  individual  is  culturally  regarded  as  of  one  sort  (e.g. 
assertive)  while  most  of  the  behaviour  of  the  other,  when  he  replies,  is 
culturally  regarded  as  of  a  sort  complementary  to  this  (e.g.  submissive). 
Cf.  p.  176. 

Configuration.   Defined,  p.  33. 

Cyclothymia.  A  syndrome  of  characteristics  of  temperament  described 
by  Kretschmer  (^Physique  and  Character,  1925).  This  temperament  is 
recognisable  {inter  alia)  by  a  tendency  to  periodic  variation  between 
gaiety  and  sadness.  It  is  believed  to  be  associated  with  pyknic 
physique.    Cf.  p.  160. 

Developmental  psychology.  The  study  of  the  changes  which  occur  in 
the  psychology  of  an  individual  as  a  result  of  growth  or  experience. 

Diachronic.  It  is  convenient  to  divide  scientific  anthropology  into  two 
major  disciplines,  the  diachronic,  which  is  concerned  with  the  pro- 
cesses of  cultural  change,  and  the  synchronic,  which  is  concerned 
with  the  working  of  cultural  systems  at  a  given  period.  Synchronic 
anthropology  ignores  the  historical  origin  of  the  details  of  culture. 
Cf.  p.  3. 

Diagonal  dualism.  A  way  of  thought  which  groups  persons  or  objects 
in  pairs  by  regarding  each  member  of  the  pair  as  the  symmetrical 
counterpart  of  the  other  (p.  239).  At  the  end  of  this  book  it  is  pointed 
out  that  "diagonal"  is  equivalent  to  "symmetrical"  in  the  sense  in 
which  the  latter  term  is  applied  to  relationships.    Cf.  p.  271. 

Direct  dualism.  A  way  of  thought  which  groups  persons  or  objects  in 
pairs,  regarding  the  members  of  each  pair  as  mutually  complementary 
(p.  239).  For  the  equivalence  between  "direct"  and  "complemen- 
tary" cf.  p.  271. 

Discrimination.   The  opposite  of  Identification  (q.v.). 

Dynamic  equilibrium.  A  state  of  aflfairs  in  a  functional  system,  such 
that,  though  no  change  is  apparent,  we  are  compelled  to  believe  that 
small  changes  are  continually  occurring  and  counteracting  each 
other.    Cf.  pp.  175  and  190. 

Eidos.   Cf.  footnote  on  p.  25;  and  pp.  30  and  220. 

Ethos.    Cf.  footnote  on  p.  2;  and  pp.  30,  32,  114,  119,  123  and  276. 

External  (or  lateral)  sanctions.  I  have  used  this  term  to  mean  those 
sanctions  which  are  imposed  upon  a  member  of  a  group  by  person^ 
outside  that  group.    Cf.  p.  98. 

Formulation.  I  have  used  this  word  loosely  as  synonym  for  cultural 
premise. 

Function.   Cf.  p.  26  et  seq. 

Grade.  The  latmul  system  of  initiatory  groups  is  not  easily  described  in 
English  and  I  have  used  the  word  "grade"  somewhat  unconvention- 
ally.   For  a  description  of  the  system  see  p.  245. 

Homicide.  A  man  who  has  killed  another.  Among  the  latmul,  the 
successful  killer  of  enemies  is  entitled  to  special  ornaments,  here 
called  "homicidal  ornaments". 

iai,  father's  mother,  father's  mother's  brother's  daughter,  all  of  the 
women  of  the  same  patrilineal  clan  as  these.  The  same  term  (or 
naisagut  or  tawonto)  may  be  applied  to  their  brothers.  For  marriage 
between  iai  and  ianan,  cf.  p.  88  et  seq. 

ianan,  son's  son  (w.s.);  father's  sister's  son's  pon  (b.s.);  the  reciprocal  of 
iai  (q.v.). 

iau,  father's  sister,  real  and  classificatory. 


Glossary  ^09 


Identification.    Cf.  explanation,  p.  35. 

Initiatory  groups.   Cf.  diagram,  p.  245. 

Institution.    Cf.  criticism  of  this  concept,  p.  27. 

Internal  sanctions.  I  have  used  this  term  to  mean  only  those  sanctions 
imposed  upon  a  member  of  a  group  by  some  other  member  (often 
an  official)  of  the  same  group,  and  I  have  not  used  it  in  the  psycho- 
logical sense  to  refer  to  conscience,  etc. 

kaishe  nampa,  a  collective  term  for  the  members  of  the  patrilineal 
group  into  which  ego's  daughter  is  married. 

kaishe-ndo,  child's  spouse's  father;  kaishe-ragwa,  child's  spouse's  mother. 
These  terms  are  also  used  between  persons  whose  offspring  might 
be  expected  to  marry,  e.g.  sister's  son's  wife  may  be  called  kaishe- 
ragwa.  The  term  kaishe  is  also  used  between  partners  who  have  ex- 
changed ceremonial  gifts  of  shell  currency. 

kamberail,  an  initiatory  group,  By  ^  or  Bx ^.   Cf.  p.  245. 

kanggat,  brother's  child  (w.s.) ;  the  reciprocal  of  iau. 

kau,  violence.    Cf.  p.  140. 

hop,  a  general  term  for  ancestral  spirits  which  guard  their  descendants 
and  help  them  in  war.    It  includes  both  mbivan  and  angk-au. 

Ian,  husband. 

lando,  sister's  husband,  sister's  son's  son,  real  or  classificatory  (m.s.). 
This  term  may  also  include  men  who  might  be  expected  to  marry 
the  speaker's  sister,  so  that  it  is  sometimes  used  instead  of  ianan. 

Ian  men  to\  "husband  thou  indeed!".    Cf.  footnote,  p.  82. 

lanoa  nampa,  literally  "  husband  people  ",  a  collective  term  for  members 
of  the  patrilineal  group  into  which  the  speaker's  sister  has  married. 
Cf.  explanation  on  p.  93. 

Lateral  sanctions.   Cf.  External  sanctions. 

laua,  sister's  child,  sister's  husband's  father  (m.s.),  and  other  relatives 
classified  with  them.    Cf.  p.  94. 

laua  nyanggu,  the  patrilineal  group  which  includes  the  speaker's  classi- 
ficatory laiias.    Cf.  p.  94. 

laua-ianan,  a  synon^tn  of  laua  nyanggu. 

Leptorrhine.  This  term  is  loosely  used  to  mean  "  having  a  narrow  nose  ". 
Cf.  p.  163. 

Logic.    Cf.  footnote  on  p.  25,  and  p.  220. 

mbapma,  literally,  a  "line",  especially  a  line  of  people  side  by  side  as 
distinct  from  a  file.  The  word  is  also  applied  to  a  group  defined  in 
terms  of  alternating  generations.  In  any  patrilineal  lineage  there  are 
tvvo  mbapma,  one  containing  the  members  of  ego's  generation,  his 
grandfather's  generation  and  his  son's  son's  generation;  the  other 
containing  members  of  his  father's  generation  and  his  son's  generation. 

mbora,  mother's  brother's  wife,  wives  of  all  men  classified  as  mother's 
brothers. 

mbuambo,  mother's  father,  mother's  mother,  mother's  brother's  son, 
mother's  brother's  son's  wife,  father's  elder  brother.  Cf.  p.  39  for 
incorrect  use  of  the  term  by  a  latmul. 

mbuandi,  a  tree  with  brilliant  orange-coloured  fruit  {Ervatamia 
aurantiacd). 

tnbwan,  the  spirit  of  an  enemy  who  has  been  killed  and  buried  under  a 
standing  stone  or  ceremonial  mound.  Such  spirits  are  regarded  as 
ancestors  and  are  believed  to  help  in  warfare  and  in  the  increase  of 
population. 

ntbtuatnggowi,  ceremonial  dolls,  which  represent  clan  ancestral  spirits 
associated  with  fertility.    Cf.  p.  233  and  Pis.  XXV  and  XXVII. 


310  Glossary 


mbwole,  a  small  ceremonial  house  for  the  boys,  especially  for  initiatory 
grades  Ax  5  and  Ay  5. 

mints hanggu,  a  mortuary  ceremony.   Cf.  p.  47. 

Moiety.  latmul  society  is  divided  into  two  totemic  groups  or  moieties 
called  nyowe  (Sun)  and  nyame  (Mother).  It  is  also  divided  into  cross- 
cutting  moieties  for  initiatory  ceremonial  (cf.  p.  245).  Membership 
in  all  these  groups  is  determined  by  patrilineal  descent. 

Motivation.  Cf.  the  examination  of  the  terms  "cognitive"  and  "afTec- 
tive"  on  p.  275. 

m.s.,  man  speaking;  placed  after  the  translation  of  a  kinship  term  it  means 
that  that  term  is  used  in  the  given  sense  only  by  a  man. 

mwai,  mythological  beings  analogous  to  wagan  (slit  gongs)  and  repre- 
sented by  masks  with  big  noses.    Cf.  p.  233  and  PI.  XXVIII  B. 

na,  father's  sister's  child  (m.s.);  daughter's  child  (b.s.);  reciprocal  of 
mbuambo. 

naisagut,  wife's  father,  wife's  mother,  wife's  brother's  wife,  wife's  mother's 
brother.    Cf.  p.  91. 

nambu  luail,  nambu  =  head,  wail  =  crocodile.  The  term  wail  is  used 
for  any  initiatory  group  (moiety,  quadrant  or  generation  group).  The 
term  "head  of  the  crocodile"  is  applied  to  the  senior  member  of  a 
generation  group  in  the  initiatory  system.    Cf.  p.  245. 

nampa,  people. 

nasa,  husband's  sister's  child;  the  reciprocal  of  mbora. 

naven,  a  set  of  ceremonial  customs  of  the  latmul  used  to  illustrate  the 
theoretical  analyses  in  this  book. 

ndjambwia,  spikes  set  in  the  ground  or  magical  devices  designed  to 
prevent  trespass.   Cf.  p.  46. 

ndo,  man. 

nemwan,  big,  great. 

nggambut,  a  post  set  up  beside  the  grave  of  a  woman.    Cf.  p.  55. 

nggambwa,  vengeance;   a  fury  who  tears  dead  bodies  to  pieces.    Cf. 

P-  57- 
nggelakaviui,  a  tuberous  epiphyte  {Myrmecodia  sp.).    Cf.  p.  59. 
ngglambi.  Cf.  p.  54  et  seq.  and  p.  140. 
nggtuail,  father's  father,  father's  father's  sister,  son's  son,  son's  daughter. 

This  term  is  also  applied  to  the  totemic  ancestors  of  a  clan. 
nggivatl-warangka.    Cf.  p.  39. 
nggiuat  keranda.  When  a  girl  goes  from  her  parents'  house  in  marriage, 

her   relatives  decorate   her.    The  ornaments  constitute  the  nggwat 

keranda,  a  ceremonial  present  to  the  husband.    Cf.  p.  105. 
nian,  son,  daughter,  and  (w.s.)  father's  sister's  child.  This  is  the  ordinary 

word  for  children  (pi.  nyanggu)  and  is  also  used  as  a  synonym  for 

nampa,  people. 
nondu,  daughter's  husband,   father's  sister's  husband,   and  husband's 

sister's  husband ;  reciprocal  of  naisagut. 
nyai' ,  father,  father's  classificatory  brother.   Cf.  mbuambo  and  tshambwi 

nyai'.  This  term  is  applied  to  father's  sister  in  naven. 
nyame,  mother,  mother's  sister  (real  and  classificatory);  any  woman  of 

mother's  clan;  son's  wife  (m.s.).  This  term  is  applied  to  mother's 

brother  in  naven. 
nyamun,  elder  sibling  of  the  same  sex,  brother  (w.s.);  father's  father. 

This  term  is  also  used  between  clans  (q.v.)  and  initiatory  grades 

(cf.  p.  245),  and  is  applied  to  elder  brother's  wife  in  naven. 
nyanggai,  sister,  real  or  classificatory  (m.s.);  father's  father's  sister  (m.s.) ; 

son's  daughter  (m.s.). 


Glossary  ^n 


Peripheral.  The  system  of  organisation  of  a  group  is  said  to  be  peripheral 
if  it  depends  for  its  sanctions  not  upon  a  higher  authority  but  upon 
the  behaviour  of  other  equivalent  groups.    Cf.  p.  99. 

Potlatch.  Competitive  ceremonial  giving,  characteristic  of  the  Indians 
of  the  North-west  Coast  of  British  Columbia. 

Pragmatic  function.  For  criticism  and  subdivision  of  this  concept, 
cf.  p.  30  et  seq.;  also  pp.  261  and  264. 

Premise.    Definition,  p.  24. 

pwivu.   Cf.  p.  46. 

rite  de  passage,  a  ceremony  accompanying  or  effecting  a  change  of  status. 

Schismogenesis.   Definition,  p.  175. 

Schizothymia.  A  syndrome  of  characteristics  of  temperament  described 
by  Kretschmer.  (Cf.  Cyclothymia.)  Schizothymia  is  characterised 
{inter  alia)  by  sudden  and  irregular  changes  from  emotional  anaes- 
thesia to  emotional  hyperaesthesia. 

Segmentation.  A  functional  system  may  be  said  to  be  segmented  when 
it  consists  of  two  or  more  parts,  each  of  which  is  in  some  degree  a  repeat 
of  the  others,  e.g.  a  system  of  clans  or  an  earthworm. 

Siblings.  Persons  of  either  sex,  who  have  one  or  both  parents  in  common. 

Social  psychology.    Cf.  tentative  definition,  p.  175. 

Social  structure.  For  the  distinction  bet\veen  this  concept  and  that  of 
cultural  structure,  cf.  p.  25. 

Sociology.  This  word  I  have  used  in  a  specially  restricted  sense.  Cf. 
pp.  30  and  34. 

Standardisation.  The  process  by  which  the  individuals  in  a  community 
are  moulded  to  resemble  each  other  in  their  behaviour.    Cf.  pp.  33, 

34.  113,  255- 
Structure.    Cf.  definitions  of  cultural  and  social  structure,  p.  25. 
Symmetrical.    A  relationship  between  two  individuals  (or  two  groups) 

is  said  to  be  symmetrical  if  each  responds  to  the  other  with  the  same 

kind  of  behaviour,  e.g.  if  each  meets  the  other  with  assertiveness. 
Synchronic.   Cf.  Diachronic,  and  p.  256. 
Syndrome.   Used  in  this  book  as  a  collective  term  for  the  characteristics 

of  a  functional  system.  This  term  is  used  in  medicine  for  the  additional 

characteristics  which  a  body  acquires  in  a  diseased  condition. 
tagail,  a  small  ceremonial  house  of  the  boys,  especially  for  members  of 

Bx^  and  By  4.    Cf.  p.  136. 
tagwa,  woman,  wife.  This  term  is  also  applied  jocularly  to  father's  mother. 
tambinyen,  a  partner  in  the  opposite  moiety,  who  is  of  the  same  genera- 
tion group  as  the  speaker,  e.g.  members  of  initiatory  group  Ax  3  are 

tambinyanggu  of  Ay  3. 
tambointsha,  feather  tassels,  a  badge  of  homicide.   Cf.  pp.  46  and  72. 
tavet,  a  rhythm  on  the  slit  gong  used  to  summon  a  man  or  group, 
tatoonfu,  literally  "wife-man",  wife's  brother,  father's  mother's  brother, 

real  or  classificatory  (m.s.).  This  term  is  the  reciprocal  of  lando  and 

is  extended  to  include  male  iai. 
timbut,  a  species  of  small  lemon  used  in  purificatory  rituals,  in  magic, 

and  as  a  purge.    Cf.  p.  155. 
towa-naisagut ,  a  collective  term  for  the  members  of  the  wife's  clan. 

Cf.  p.  93- 
Transvesticism.   The  wearing  of  women's  clothes  by  men,  or  of  men's 

clothes  by  women. 
tshaishi,  wife  of  elder  brother,  real  or  classificatory  (m.s.). 
tshambivi  nyai',  own  father's  younger  brother. 
tshat  kundi,  the  act  of  addressing  the  sister's  child  as  an  ancestor.  Cf .  p.  44. 


312  Glossary 


tshitnangka,  a  species  of  fish ;  an  act  in  initiation  in  which  an  initiator 

dances  like  a  fish. 
tshimbtuora,  a  fringe  of  hanging  strips  of  pahn  leaf  used  as  a  screen  in 

initiation  ceremonies.  This  word  is  used  by  wagan  (shamans)  for 

water.    Cf.  p.  47. 
tshivla,  long. 
tshuambo,  younger  sibling  of  the  same  sex;  son's  son  (m.s.);  husband's 

younger  brother. 
Tshuosh,  a  tribe  who  live  immediately  north  of  the  latmul. 
tshugukepma,  ceremonies  in  which  the  sister's  children  dance  as  the 

ancestors  of  their  mother's  brother's  clan.   Cf.  p.  10. 
tshumbuk,  a  pointing  stick  used  in  sorcery.  This  object  is  personified. 

Cf.  pp.  60  and  72. 
wagan.  This  term  is  used  among  the  Central  latmul  in  two  senses:  (a)  for 
jA       certain   clan  ancestral  spirits  who  "possess"   shamans  and  speak 

through  their  mouths ;  (6)  for  the  sacred  slit  gongs  used  in  initiation. 

Among  the  Eastern  latmul  the  term  is  used  only  for  the  gongs  and 

a  separate  term  {lemwail)  is  used  for  the  spirits.  The  Central  latmul, 

however,  regard  the  gongs  and  spirits  as  fundamentally  the  same. 

Cf.  pp.  55,  233  and  236. 
waingga,  purchase.   This  term   is  also  applied  to  the  acquisition  of  a 

wife.    Cf.  p.  79. 
tvarangka,  father's  father's  father. 
tvau,  mother's  brother,  son's  wife's  brother,  and  other  relatives  classified 

with  them.    Cf.  p.  94. 
wau-mbuambo,  a  collective  term  for  the  members  of  the  clan  of  own  or 

classificatory  mother. 
4  ivindjimbu,  spirits  which  live  in  trees.    Cf.  p.  65. 
w.s.   Woman  speaking;  placed  after  the  translation  of  a  kinship  term,  it 

means  that  that  term  is  used  in  that  sense  only  by  a  woman. 
yigen,  beautiful,  gentle,  quiet.    Cf.  p.  141. 
yigen  kundi,  literally  "  quiet  singing  ".  A  mortuary  ceremony  performed 

by  women.   Cf.  p.  156. 
yivut.   Cf.  footnote,  p.  129.  \    - 


4 


Xj^^ 


^^S- 


y' 


V 


PLATES 


The  dancing  ground  of  Palimbai  in  the  flooded  season.  The  wompunau  or  dancing 
ground  is  a  long  avenue  down  the  centre  of  the  viHage.  In  it  stand  the  men's 
ceremonial  houses,  of  which  one  is  visible  in  the  picture.  The  sides  of  the  wom- 
punau are  lined  with  mounds  on  which  coconut  palms,  crotons,  bananas,  etc.  are 
planted  to  raise  them  above  the  flood  level.  Dwelling  houses  lie  on  both  sides  of 
the  wompunau,  hidden  by  this  vegetation.  The  photograph  was  taken  from  the 
upper  storey  of  another  ceremonial  house.  In  the  immediate  foreground  are  the 
tops  of  Dracaena  plants  growing  on  the  wak  or  ceremonial  mound  belonging  to 
this  house.  Behind  these  a  woman  is  paddling  a  small  canoe  across  the  wompunau. 
In  front  and  to  the  left  of  the  far  ceremonial  house,  its  wak  is  xisible. 


PLATE  I 


A.  \uven  in  Paliinbai  for  a  sister's  son  who  had  made  a  large  new  canoe.  The 
two  youno;  men  are  ivaiis  ( mother's  brothers )  and  are  dressed  as  old  women,  with 
tattered  raineapes  and  bedraggled  skirts.  They  hobble  about  the  \illage  support- 
ing themselves  by  means  of  short  paddles,  such  as  are  used  by  women  in  canoeing. 
(Men  use  paddles  10  feet  to  12  feet  long  and  stand  while  paddling. )  The  nearer 
of  the  two  trans  carries  a  white  fowl  in  his  hand  to  be  presented  to  his  laua. 


H.   Mavcu  in  Palniibai:  tlic  two  tcaus  searching  for  their  laua 

PLATE  II 


A.  Navcn  in  Falimbai.  The  icaiis  iiie  on  the  way  to  tlic  canoe 
made  by  their  laua.  Suspended  from  the  nose  of  the  further 
wau  is  seen  an  "ornament"  made  of  old  sago. 


B.   Nuven  in  Pahi 


ihai :   one  o: 


f  the  waus 


PLATE  III 


A.  Naven  in  Palimbai.   One  oi  the  wans  stumbling  because  of  his  pretended  feeble- 
ness.   The  children  crowd  round  laughing. 


fel.' 


B.   Naven  in  l-'alinibai.    C^ne  ot  the  nans  has  readied  the  canoe  and  tails  uito  it  witli 
his  legs  widespread.   The  other  is  arri\  ing  miserably  on  all  fours. 


PLATE  I\' 


A.  A  small  naven  in  Kankanamun. 
The  laua  luinipa  (sister's  children)  had 
assisted  in  building  operations.  When 
the  work  was  done,  one  of  the  waui 
dressed  himself  as  a  woman  and  is  here 
seen  dancing  with  a  head  taken  from 
the  enemy  by  one  of  his  /«»«s.  The 
head,  a  portrait  modelled  in  clay  on  an 
enemy's  skull,  is  an  old  one. 


B.  Initiation  in  Kankana- 
mun. The  novice  is  sepa- 
rated from  his  mother,  and 
his  wau  ceremonially  takes 
o\'er  her  function  and  is 
called  mjume  ("mother"). 
The  tcdii  is  seen  carrying 
a  novice  pick-a-back  as  a 
real  mother  might  carry  her 
small  child. 


PLATE  V 


A.  Women's  noven  in  Kankananiun.  A  !)()>  luis  made  a  new  canoe,  and  the  photo- 
i^rapli  shows  his  "mothers",  father's  sisters,  and  elder  brothers'  wives  celebrating 
this  achie\ement.  This  was  only  a  small  naven  and  the  transvesticism  is  of  a 
makeshift  \ariety.  The  two  "mothers"  have  discarded  their  skirts  and  sit  in  the 
canoe  using  women's  paddles,  i.e.  they  are  "naked"  and  not  transvestite.  The 
fatliors'  sisters  and  elder  brothers'  wives,  although  the>-  are  using  women's 
paddles,  stand  to  paddle  in  the  manner  of  men. 


B.  The  same  navcn.  The  pliotograph  shows  the  war-like  gestures  of  the 
transvestite  women.  One  of  tlicm  is  wearing  a  pubic  apron  made  of  lea\'es, 
a  makeshift  cop\  ot  tliat  worn  by  men. 


PLATE  \T 


A.  Men's  ceremonial  liouse  in  Kan- 
kanaiiinn.  Tliis  was  the  finest  eere- 
nionial  house  in  any  of  the  river 
N'illatjes.  The  sliinyles  of  the  sauo- 
leaf  tliuteh  are  alternately  yellow 
and  brown,  gi\ing  a  deeoratixe  pat- 
tern. In  the  gable  front  f.nir  little 
windows  are  visible,  with  an  en- 
emy's skull  in  eaeh.  Above  these 
windows  there  is  a  grotesque  face 
of  which  only  the  teeth  and  cres- 
cent-shaped nose  ornaments  are  \  is- 
ible  in  the  photograph.  In  front  of 
the  building  is  the  wak  or  cerenu)- 
nial  mound  on  which  dead  enem> 
bodies  and  captives  are  laid.  At 
tlie  foot  of  the  ladder  is  a  little  tree 
of  the  species  called  wani,  a  totemic 
ancestor  of  the  clan  which  owns 
this  end  of  the  ceremonial  house. 
The  building  is  130  feet  in  length. 


li.  A  man  debating.  The  man  was 
posed  for  this  photograph.  He 
stands  beside  the  elaborately  car\  ed 
debating  stool,  and  holds  in  his 
hand  the  three  bundles  of  leaves 
with  which  he  beats  the  stool  dur- 
ing his  speech.  Photograph  taken 
in  a  ceremonial  house  in  Malingai 
village. 


PLATE   \  11 


A.  The  coiiKi  (il  a  LxiLiiiDiiial  lioii^t-  in  ralinihai.  The  photo.yraph  shows  one  of 
the  posts  whieh  support  the  floor.  The  top  of  the  post  is  earved  to  represent  a 
iciudjimlju  or  wood  spirit. 

u.  Uaisinjf  the  ridge  pole  for  one  of  the  ceremonial  houses  in  Palimbai.  A  scaf- 
folding has  been  constructed  around  the  supporting  post.  The  horizontal  poles 
in  this  scaffolding  are  not  tied  do\\'n,  so  that  they  can  be  remo\ed  to  allow  the 
ridge  pole  to  be  raised.  The  horizontals  are  replaced  under  it  at  each  step.  The 
ridge  pole  is  raised  by  men  beside  it  on  the  scaffolding,  and  by  others  below  who 
push  on  a  cross-piece  tied  to  a  bamboo  shaft.  The  shaft  is  attached  to  the  ridge 
pole  at  the  top  by  a  loop  of  cane.  Such  a  work  is  the  duty  of  all  the  affinal  relati\es 
of  the  clan  whose  ceremonial  house  is  being  constructed.  The  work  would  be 
accompanied  b\'  beating  of  gongs  and  pla\ing  of  flutes. 


PLATE  \  III 


A.  A  pair  of  slit  jiongs  in  Mindinibit.  Vhe  photoj^raph  shows  only 
the  elaborately  carved  ends  of  the  gonj^s.  The  carving  on  both  these 
specimens  represents  the  heads  of  prawns  with  their  rostra.  The 
interest  in  noses  gives  a  special  importance  to  such  totemic  ancestors 
as  possess  conspicuous  noses,  e.g.  sago  weevils,  prawns,  etc.  On  the 
nearer  gong,  at  the  base  of  the  rostrum,  a  frog  is  represented  with  its 
hind  leg  extended  in  a  long  wavy  line. 


a.  Initiation  in  Kankanamun.  The  initiators  are  waiting  in  two  lines  armed  with  sticks. 
The  no\ice  will  enter  through  the  leaf  screen  at  the  back  of  the  picture  and  will  nm  the 
gauntlet  of  the  initiators.  Actually,  the  no\'ice's  father  will  come  with  him  and  will 
protect  the  back  of  the  noxice,  allow  ing  blows  to  fall  on  his  own  back.  This  beating  of 
the  father  is  analogous  to  a  beating  of  the  novice,  since  both  are  of  the  same  initiatory 
moietv. 


PLATE  IX 


Initiation  in  Malinj^ai.  The  novice  is  lying  prone  on  an  inverted  canoe  claspii 
his  mother's  brother  who  acts  as  a  comforter  and  "mother".  An  initiator  of  tl 
opposite  moiety  to  the  no\ice  is  cutting  the  latter 's  back  with  a  small  bamb( 
blade.  In  the  foreground  is  a  bowl  of  water  with  swabs  of  fibre  to  wipe  away  tl 
blood.  The  white  and  black  paint  on  the  faces  of  the  two  men  is  a  privilege 
those  who  have  killed  a  man,  and  is  worn  on  all  ceremonial  occasions.  The  bai 
of  opossmn  fur  worn  b\'  the  initiator  is  also  a  badge  of  homicide. 


PL.VTE  X 


^^■' 


A.  Initiation  in  Kankananiun.  The  novice  is  sitting  in  the  lap  of  Ins  mother's  brother 
while  initiators  cut  crescents  above  the  novice's  nipples.  The  initiator  with  bird-of- 
paradise  feathers  in  his  hair  is  doing  the  cutting,  while  others  observe  his  craftsman- 
ship and  hold  the  novice. 


B.   The  same  initiation:  the  cuttiTig  of  tlie  uomcc  s  iiatk.    Otiicr  initiators  watcli  the 
operation. 

PLATE  XI 


Initiation  in  Koniindimhit  \illatie:  Inillying  the  novice.  For  ai:)Out  a  week  after  his 
back  lias  been  cnt,  the  novice  is  subjected  every  morning  to  a  series  of  bullying 
ceremonials.  He  is  made  to  squat  like  a  woman  while  masked  initiators  maltreat 
him  in  various  ways.  The  particular  incident  shown  here  is  a  recent  innovation. 
The  mask  worn  by  the  initiator  was  made  by  people  in  the  mountains  at  the  head 
of  the  Yiuit  River.  This  mask  somehow— probably  as  loot— came  into  the  possession 
of  L.ower  Sepik  natives.  A  party  of  returned  latmul  labourers  stole  the  mask  on 
their  way  to  Komindimbit,  and  it  was  adopted  into  the  local  initiation  ceremonies; 
it  was  referred  to  as  tumhiuni  ( the  ordinary  pidgin  word  for  a  mask ) .  The  tumhuan 
divines  for  theft.  The  cow's  bone  ( acquired  on  some  plantation )  which  is  on  the 
ground  in  front  of  the  figure's  knees  is  held  up  on  a  string  in  front  of  the  novice's 
face.  The  tumhuan  then  says,  addressing  the  novice  in  the  feminine  second  person 
singular:  "Are  you  a  child  that  steals  calico?"  He  then  moves  his  arm  causing  the 
bone  to  swing,  which  indicates  an  affirmative  answer  to  the  divination;  and  he  slaps 
the  no\ice.  The  same  question  is  asked  about  yams,  bananas,  tobacco,  etc.,  etc. 
In  each  case  an  affirmative  answer  is  olitained  and  the  no\ice  is  slapped. 


PLATE  XII 


Initiation  in  Komindimbit:  an 
initiator  expressing  scorn  for  a  nov- 
ice by  rubbing  liis  buttocks  on  the 
latter's  head.  This  gesture  is  pre- 
sumabh-  quite  distinct  from  the 
gesture  of  the  icaii,  who  shames 
himself  b>-  rubbing  his  buttocks  on 
his  laud's  leg.  ( The  dog  in  this  pic- 
ture belonged  to  the  ethnographer 
and  was  not  nati\e. ) 


B.  Initiation  in  Komindimbit:  a 
no\ice  eating  a  meal.  He  is  made 
to  wash  before  he  eats,  and  even 
then  must  not  touch  his  food  with 
his  hands,  but  must  pick  it  up  awk- 
wardh-  and  miserably  with  a  pair  of 
bamboo  tongs  or  a  folded  leaf.  This 
was  tile  youngest  of  the  novices  in- 
itiated in  this  ceremony.  The  parts 
of  his  bod\-  which  ha\e  been  cut  are 
wet  with  oil  and  the  remainder  of 
Ills  body  is  smeared  with  cla>-. 


PLATE  XIII 


Initiation  in  Koniineliniljit.  Evci)  cla>  tor  noininalK'  five  days  bnt  actually  for  a 
rather  longer  period,  the  novices  were  bullied  during  the  morning  and  allowed  to 
rest  in  the  afternoon.  This  photograph  shows  them  resting,  sitting  or  lying  on  mats 
on  the  ground  in  the  ceremonial  house.  The  initiators  sit  on  raised  platforms  above 
the  novices  whom  they  are  initiating.  Down  the  centre  of  the  ceremonial  house  the 
slit  gongs  are  arranged  in  pairs,  and  initiators  of  the  opposite  moiety  are  sitting  on 
the  platforms  on  the  far  side  of  the  house.  Below  them  on  the  groiuid,  hidden  by 
the  gongs,  are  the  novices  whom  they  are  initiating.  A  pair  of  flutes  is  lying  on  the 
central  gongs,  and  in  front  of  the  nearest  house-post  there  is  a  pair  of  water  drums. 
Feather  headdresses,  etc.,  belonging  to  the  initiators  are  suspended  on  hooks  from 
the  beams. 


PLATE  XIV 


A.   Dwelling  house  in  Kankanannm  village 


B.  Interior  of  house  in  Kankanamun  \  illage.  The  foreground  is  filled  by  mosciuito 
bags  made  of  plaited  rushes.  Usually,  a  husband  has  his  own  bag  and  each  wife 
has  a  bag.  A  separate  larger  bag,  e.g.  that  on  the  extreme  right  of  the  photograph, 
is  pro\ided  for  the  children,  who  sleep  together.  Suckling  children  sleep  with  their 
mothers.  Suspended  in  the  house  are  a  number  of  baskets  and  patterned  string 
bags  imported  from  the  Tshuosh  people.  The  camera  was  set  up  on  the  floor  in  the 
clear  space  at  the  front  end  of  the  house,  and  there  is  a  corresponding  clear  space 
beyond  the  mosquito  bags  at  the  far  end. 


PLATE  X\^ 


A.  WoiiK'ii  of  Mindinibit  goin<4  aloiiu  tlic  lianks  of  the  Scpik  in  small  canoes  to  their 
fish  traps 


H.  Women  of  Pahmbai  fishing  in  a  lake.  The  circular  net  is 
held  by  two  women  who  stand  in  the  prows  of  two  small  canoes. 
Each  canoe  is  propelled  by  a  woman  sittin,<j  in  the  stern.  Such 
a  net,  called  cliiira,  is  used  to  catcli  the  fish  which  hide  under 
small  islands  of  floating  grass.  The  net  is  pushed  under  the  edge 
of  the  island  and  then  raised.  A  small  patch  of  grass  is  some- 
times simpl\-  lifted  up  in  the  net  and  tlien  thrown  out. 


PLATE  \\l 


A.  \\'()inan  of  Miiulimhit  \il- 
lage  with  child,  the  hitter 
h  e  a  \-  i  1  y  ornamented  with 
shells,  as  is  usual  even  in 
everyday  life.  Both  woman 
and  child  are  rather  shy  of 
the  ethnographer. 


H.  Woman  ot  Kankanamun 
\illage  with  child.  She  did 
not  know  that  she  was  being 
photographed. 


PLATE   X\'II 


?T'r* 


A.  An  audience  «f  women  and  children  \vatcliin<j;  a  spectacle  staged  l:»y  men.  Tlie 
voung  women  and  especially  the  children  are  hea\ily  ornamented  for  the  occasion. 
The  spectacle  in  this  case  was  a  xvcigcm  dance. 


H.  Old  men  of  age  grades  By  2  and  B.v  2  dancing  as  tvagan.  Wagan  are  spiritual 
beings  mythologically  connected  with  water.  Their  voice  is  represented  by  the 
beating  of  slit  gongs,  a  secret  of  late  initiation.  The  men  impersonating  tcagan  are 
here  seen  prodigiously  decorated,  dancing  with  a  djuni  (fish  net,  cf.  Plate  X\T  b) 
in  which  the  xcagan  were  raised  from  water  in  their  mythological  origin.  The  net 
contains  an  imitation  fish  made  of  wood  which  is  tied  to  the  bottom  of  it. 


PLATE  X\  III 


A.  Woman  painted  and  ornamented  for  tslnigii- 
kepma  procession,  a  public  ceremonial  in  which 
the  women  march  in  a  column  up  and  down  the 
dancing  ground.  They  celebrate  the  ancestors  of 
their  icau-mjame-numpu  (maternal  clan).  For 
such  an  occasion  the  women  wear,  in  addition  to 
their  best  skirts  and  other  ornaments,  a  number 
of  decorations  which  are  usually  only  worn  by 
men.  In  the  photograph  the  headdress  of  casso- 
wary skin  and  the  bobbing  fans  set  with  white 
fowls'  feathers,  are  both  of  them  normally  worn 
by  men,  as  also  are  tlie  shell  rings  which  the 
woman  is  wearing  on  her  arm.  The  paint  on  her 
face  is  yellow  ochre  and  not  the  white  paint  worn 
by  homicides. 

B.  In  front  of  the  tslui^u-kcpma  procession  (cf. 
.supra )  a  group  of  women  perform  a  dance  rep- 
resenting an  incident  in  the  mythology  of  the 
maternal  clan.  The  figure  in  the  left  front  of 
the  pictiue  is  a  woman  wearing  a  faceless  mask 
shaped  to  represent  the  Aibom  mountain.  Part 
of  a  second  mask  representing  the  Tchambuli 
mountain  is  visible  on  the  extreme  right.  These 
two  mountains  are  a  conspicuous  feature  of  the 
Sepik  fens.  All  pieces  of  land  are  m\  thologically 
supposed  to  be  floating  islands,  and  here  we  see 
a  woman  steering  the  floating  mountain  with  a 
long  paddle  such  as  is  normally  used  by  men. 
Her  gesture,  too,  is  reminiscent  of  male  pride. 


PLATE  XIX 


A.  A  ureat  man  oF  Falimhai  had  died  the  niglit  before  and  \\  as  buried  in  the  earl\-  morning.  Later 
in  the  morning,  tlie  figure  here  shown  was  constructed  to  represent  him.  The  head  of  the  figure 
\\  as  an  unripe  coconut  and  the  body  was  made  of  bundles  of  pahii  leaves.  Spears  were  set  up 
against  it  with  their  points  stuck  in  the  figure  to  mark  the  places  where  he  had  been  wounded  in 
fight,  and  other  spears  were  stuck  into  the  ground  beside  the  figure  for  those  which  he  had 
dodged.  A  series  of  \ertical  spears  of  which  only  two  are  \isible  in  the  photograph  was  set  up 
in  front  of  the  figure  according  to  his  achievements.  The  figure  itself  was  ornamented  with  shells, 
etc.  Si.x  sago  baskets  were  suspended  from  its  right  shoulder  to  represent  his  si.\  wi\'es,  a  string 
bag  was  suspended  on  the  left  shoulder  representing  his  skill  in  magic.  A  number  of  sprigs  of 
ginger  in  its  headdress  represented  persons  whom  he  had  invited  to  the  \illage  so  that  otlier 
people  could  kill  them.  In  the  right  hand  of  the  figure  (not  visible  in  the  photograph)  was  a 
dr\-  lump  of  sago— since  it  was  said  that  in  his  lifetime  he  had  once  killed  a  bird  b>-  throwing  a 
lump  of  sago  at  it.  The  branch  of  timhut  (lemon)  set  in  the  ground  beside  the  figure  is  symbolic 
of  his  knowledge  of  m\  thology. 

B.  The  figure  was  set  up  l)y  the  initiatory  moiety  of  which  the  dead  man  had  been  a  member. 
Later,  members  of  the  other  moiety  came  and  took  away  the  \  arious  symbols  one  by  one.  In  this 
photograph  a  man  is  seen  taking  away  one  of  the  spears  denoting  a  wound,  claiming  that  he  has 
liad  a  similar  wound.  On  the  ground  by  the  furtlier  foot  of  the  figure  is  a  broom  and  a  pair  of  the 
boards  used  for  picking  up  rubbish.  These  objects  are  symbolic  of  tlie  work  whicli  the  dead 
man  had  done  in  cleaning  the  ceremonial  house  during  his  lifetime. 


PLATE  X.X 


A.  The  row  of  spears  in  front  of  the  mortuary  figure  shown  in  Plate  XX.  Tlie  figure 
is  seen  in  the  haekground.  In  front  of  it  there  are  (a)  eight  spears  for  men  lie  had 
killed,  ( b )  ten  spears  for  men  killed  by  other  members  of  the  eanoe  on  whieh  he 
had  ser\ed  in  the  prow,  (c)  ten  spears  for  wild  pigs  whieh  he  had  killed:  these 
are  marked  with  pieces  of  banana  leaf,  and  ( (/ )  ( not  \'isible  in  the  photograph ) 
si^ears  for  the  eroeodiles  and  flying  foxes  whieh  he  liad  killed. 


B.  Figures  in  mintshonggu  mor- 
tuary ceremony,  Palimbai.  The 
skulls  of  the  dead  ha\'e  l^een 
cleaned  and  portraits  modelled 
upon  them  in  clay.  These  are 
set  up  as  the  heads  of  the  figures, 
which  stand  on  a  platform  sus- 
pended from  the  roof  of  the 
house.  Both  the  figures  and 
platform  are  prodigiously  orna- 
mented. In  front  of  the  two  fig- 
ures is  a  portrait  skull  of  the  man 
whose  mortuary  figure  is  shown 
in  Plate  XX.  The  ceremony 
takes  place  at  night,  when  the 
ancestral  songs  of  the  deceased 
are  sung  and  men  concealed 
under  the  platform  play  on  the 
clan  flutes. 


PLATE  XXI 


Man  of  Mindiinbit  \  illa>j:r  ol  llie  proud,  excitable,  dramatisinj^  personality  which  is 
preferred  in  men.  Tliis  man  was  indeed  considered  by  the  latmid  to  be  somewhat 
too  unstable.  He  was  an  enthusiastic,  but  confvised  and  unreliable,  informant.  His 
posture  shows  the  male  reaction  to  the  camera  ( cf .  Plate  XXVI ) .  His  hand  grasps 
his  lime  stick,  ready  to  scrape  it  on  the  mouth  of  the  lime  box,  to  produce  a  loud 
noise  expressive  of  pride  or  an<j;er. 


PLATE  XXII 


A.  Mali-kiiKljiu  oi  Kiuikananiuii,  whose  personal- 
ity is  described  in  the  chapter  on  "Preferred 
Types" 


B.  A  man  of  Tanil)unum  village 


PLATE  XXII I 


A.  Nati\c  of  Mindimbit.  A 
man  of  the  distreet  type  and 
slijihth*  pyknic  physitiue:  a  ver\ 
intellijient,  humorous  man  and 
one  of  my  most  knowledgeable 
informants,  to  whom  much  eso- 
ti'ric  knowledge  had  been  en- 
trusted because  his  elders  had 
no  fear  of  his  declaiming  their 
secrets  in  a  sudden  fit  of  temper. 
The  pendants  on  his  lime  stick 
are  a  tallv  of  tlie  men  he  has 
killed. 


li.  'rsliiiiihat,  a  uatixc  of  Kanka- 
nainun,  a  product  of  culture  con- 
tact 


ftiMTiim 


PLATE  XXIV 


11a  poi trait  skull  at  a  woman.  Slir  was  a  native  of  Kankanammi  who  died  some 
three  generations  ago.  Her  skidl  was  eleaned,  exhibited  at  mortuary  eeremonies,  and 
finally  buried  as  is  customary.  But  as  she  was  considered  to  be  strikingly  beautiful, 
the  men  later  dug  up  her  portrait  ( and  probably  substituted  another  skull  in  the 
grave ) .  Since  then,  her  skull  has  been  used  in  mhwatnggowi  ceremonies  ( cf.  Plate 
XXVII ) .  Her  long  nose  was  especially  admired.  In  the  photograph,  the  breast  is  a 
half  coconut  shell. 


PLATE  XX\' 


A  and  H.  Two  women 
of  Kankananiun  sliow- 
injf  niodesh'  before  the 
c  a  m  era  ( cf .  Plate 
XXII ) 


platp:  x.w'i 


A  mbwatnggowi  figure,  Kankanamiin.  A  portrait  skull  of  a  dead  man,  chosen  for 
his  good  looks,  is  set  up  as  the  head  of  a  doll  and  prodigiously  ornamented.  The 
white  object  above  the  head  is  a  European  plate  which,  being  brighter  and  whiter, 
is  used  instead  of  a  Melo  shell.  The  shell  ornament  is  hiuig  as  an  extension  to  the 
nose.  The  ceremony  is  an  initiatory  secret,  but  is  also  said  by  a  few  men  to  promote 
the  prosperity  and  fertility  of  the  xillage.  The  dance  takes  place  at  night  behind  a 
screen  of  totemic  banana  leaves.  The  men  lift  up  the  dolls  by  the  poles  on  which 
they  are  constructed,  raising  them  high  abo\e  the  screen  so  that  the  women,  stand- 
ing as  an  audience,  see  the  figures  as  Jack-in-the-boxes.  Each  clan  has  its 
mbwatnggowi,  and  the  principal  interest  of  the  natives  in  the  ceremony  lies  in  its 
spectacular  nature  and  its  esoteric  origins  rather  than  in  its  eflect  upon  fertility. 


PLATE  XXVII 


A.  Figures  of  wa^aii  with  enormous  noses,  in  Jentschan.  This  photo- 
graph shows  the  setting  of  the  dancing  ground  for  icagan  dances.  In 
tlie  liackground  to  the  right,  the  front  of  the  ceremonial  house  is  visible. 
In  front  of  this  is  a  screen  of  banana  leaves  through  which  the  dancers 
impersonating  wagan  will  come.  The  two  figures  in  front  of  this  screen 
have  been  hastily  and  secretly  made  ready  and  ornamented  during  the 
two  days  preceding  the  dance.  They  were  set  up  in  the  dance  ground 
in  the  small  hours  of  the  morning,  so  that  the  women  woke  to  find  the 
dancing  ground  prepared.  Each  figure  consists  of  a  post  set  at  a  high 
angle  in  the  ground;  to  the  top  of  this  post  a  face  is  attached.  The  face 
is  an  enormous  loop— the  nose— near  the  base  of  which  two  half  coconut 
shells  are  fixed  to  represent  eyes.  From  the  junction  of  the  face  and  post 
a  sago  frond  is  suspended,  and  slopes  forward  on  to  the  ground.  The 
blades  of  the  leaflets  of  this  frond  have  been  remo\ed,  and  on  to  the 
mid-ribs  of  the  leaflets  large  orange-coloured  fniits  ha\e  been  threaded. 
This  sago  leaf  is  described  as  the  ioli  ( raincape )  of  the  figure.  The 
post  itself  is  entirely  co\'ered  with  totemic  ancestral  plants,  one  of  the 
figures  being  decorated  with  ancestors  of  the  Mother  moiety,  the  other 
with  ancestors  of  the  Sun  moiety.  In  front  of  the  two  figures,  a  rectangle 
is  enclosed  with  white  palm  fronds.  This  is  the  "lake"  in  which  the 
tcagdu  w  ill  dance. 

B.  The  head  of  a  mwai  mask,  Kankanamun.  Mwai  are  extinguisher- 
shaped  masks,  and  the  ceremony  in  which  they  perform  is  held  in  the 
dancing  groimd  of  the  ta<iuil  ( jimior  ceremonial  house ) .  The  ceremony 
is  a  junior  analogue  of  the  icagan  dances.  The  photograph  shows  the 
carved  face  detached  from  the  extinguisher-shaped  framework.  The 
nose  is  extended  into  a  long  appendage  ending  in  a  snake's  head.  The 
face  is  made  of  wood,  to  which  shells  are  attached  by  means  of  cfay 
worked  with  lime  and  oil.  The  face  is  hanging  in  a  dwelling  house  witli 
\ari()us  shell  ornaments  in  the  background. 


'v.y'- 


PLATE  XXVIII 


^ 


vJ3 

C 

3JI 


University  of  California  Library 

'"!il«;  lx)ok  :>  OVL  on  the  list  d.uc  sianiped  L»tiow. 


i-i^ 


'% 


^ 


\ 


-< 


-< 


>^ 


J-C/IV 

(JLOCTl  91998 

NOV  0  1  mx 
A  Pp. 


1995 


\i 


OS. 


)SAfJCFlfj> 


ler 


,.  m 


!? 

'  -'J  ljj:i\  MJI"- 

% 

>s>^,fl!BRARY(9/r^ 

t—' 

r^ 

i<j 

OO 

i^                                       ""^ 

j^ 

SO 

— -                                         ■" 

> 

•^ 

39 

-< 

V^ 

^.!/0JI"IV3J0>^ 

'< 

.'dU-i^'^' 


S  </. 


'^n\Ymm'^     '^^imwm^'^ 


^^AHvagiii^"^        <rii30NV.soi>^     %„..,.u^vi     i 


3         <£ 

P       o 


niii  III  III!  II  iiiiii   III  nil  ml  III  ,1'  M  III  II  iiiii 

3  1158  00119  607 


nr^^ 


^UIBRARYO^         ^IBRARY^/ 

ilKTi  §} 


\\\l  l'M\ ' 


'>-       i^'- 


3  C^-     -liJ 


"^AaMINQJUV^ 


^lOSAfJCEtfj>- 


%il3AIN(1]Wl 


^ 


^OFCAIIFO/?^ 

IN 


^' 


>         £^ 


vVW^ 


P        c£ 


3        O 


3       O 


'^Aa3AINfl 


^ILIBRA[ 


'%dllV] 


Gi^^ 


W87 


^^ 


ft]  05 1 C 

DU 


— iii^ 


O